JP4218283B2 - Target projection type three-dimensional shape measuring method and apparatus - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a target projection type three-dimensional shape measuring method and apparatus for accurately measuring a three-dimensional shape of a huge target object of several m square or more in a short time.
[0002]
[Prior art]
JP-A-3-131706, JP-A-5-26639, JP-A-2001-116526, etc. have already been proposed as means for measuring the three-dimensional shape of a huge target object of several m square or more.
[0003]
As shown in FIG. 10, a “three-dimensional position measurement device” disclosed in Japanese Patent Laid-Open No. 3-131706 has first and second imaging devices 52 and 53 that image an object 51. Is a device for measuring the three-dimensional position of an object based on two images obtained from the above. In addition, a slit light projector 55 that projects the slit light 54 toward the space including the object 51 is provided, and the depth of the slit image on the object surface is obtained based on the principle of triangulation, and binocular stereoscopic vision is obtained. Thus, the search range on the horizontal scanning line in the correspondence search is limited.
[0004]
In addition, as shown in FIG. 11, the “three-dimensional measurement method” disclosed in Japanese Patent Application Laid-Open No. 5-26639 is provided with reference point members 62 at least four places around the workpiece 61 or on the object to be measured. A reference bar 63 having a known distance is attached, and when calculating the position of the measurement point based on image data from two or more CCD cameras 64, first, the reference point member and reference bar 2 in the image data are calculated. The absolute position of two or more CCD cameras is calculated from the distance of the points, and then the position of the measurement point is calculated based on this data. Further, every time the image data changes, two or more CCD cameras are calculated. The absolute position of the CCD camera is recalculated.
[0005]
In addition, as shown in FIG. 12, the “three-dimensional shape measuring apparatus” disclosed in Japanese Patent Application Laid-Open No. 2001-116526 generates a pattern light composed of a plurality of regions whose intensity is modulated to three or more values based on a pattern code using a laser beam. Projecting means 71 for projecting onto an object, photographing means 72 for obtaining a pattern image by photographing light reflected from the object by the projection of the pattern light, and calculating distance information to the object based on the pattern code and the pattern image And calculating means 73.
[0006]
[Problems to be solved by the invention]
Conventional three-dimensional shape measuring means includes means for measuring a three-dimensional shape of an object by projecting a position detection target onto the object (for example, Japanese Patent Application Laid-Open No. 2001-116526), and a light cutting method using laser slit light (Japanese Patent Application Laid-Open No. Hei 3). -131706), means for attaching a reference bar (Japanese Patent Laid-Open No. 5-26639), means for calculating the projection position of the spot light by projecting the spot light and photographing from a plurality of viewpoints, and the like.
[0007]
However, the detection points are limited in the position detection of the reference bar and spot light. Further, when a position detection target is projected onto an object, it takes a long time to detect many points. Further, the light cutting method using laser slit light has a problem that the detection position cannot be strictly narrowed down and the accuracy is insufficient.
[0008]
The present invention has been developed to solve the above-described problems. That is, an object of the present invention is to provide a three-dimensional shape measurement method and apparatus capable of accurately measuring a three-dimensional shape of a huge target object of several m square or more in a short time.
[0009]
[Means for Solving the Problems]
  According to the present invention, two vertical and horizontal slit lights are sequentially applied to an object while changing positions.SeparatelyIrradiate the projection position of each slit lightOne unit from different positionsphotographAt the same time, a reference scale serving as a dimensional reference is attached to the object, the reference scale is imaged, and based on each image of the slit light taken, each intersection, vanishing point, and reference scale of the vertical and horizontal slit light are two-dimensional. The position is calculated for each of the different shooting positions, and the object is calculated from the obtained two-dimensional position data.A target projection type three-dimensional shape measuring method characterized by measuring a three-dimensional position is provided.
[0010]
  Further, according to the present invention, the vertical and horizontal slit lights are sequentially applied to the object while changing the position.SeparatelyThe slit light irradiation device to irradiate and the projection position of each slit light from different positionsIn one unitShootOne or moreA digital camera,A reference scale that is affixed to an object and serves as a dimensional reference, and the digital cameraTakenBased on each image of the slit light, the two-dimensional positions of the intersections, vanishing points and reference scales of the vertical and horizontal slit lights are calculated for each of the different photographing positions, and the object is obtained from the obtained two-dimensional position data.Calculate 3D positionArithmetic unitAnd a target projection type three-dimensional shape measuring apparatus.
[0011]
  According to the method and apparatus of the present invention described above, the slit light irradiating device applies the two vertical and horizontal slit lights to the object while sequentially changing the position.SeparatelyIrradiate and project the projection position of each slit light with a digital cameraOne unit from different positionsphotographAt the same time, a reference scale as a dimensional reference is affixed to the object, and this reference scale is imaged.Depending on the arithmetic unit,Based on each image of the captured slit light, the two-dimensional position of each intersection, vanishing point and reference scale of the vertical and horizontal slit light is calculated for each of the different photographing positions, and the object is obtained from the obtained two-dimensional position data. ofBy calculating the three-dimensional position, the three-dimensional shape of a huge target object of several m square or more can be accurately measured in a short time.
[0012]
According to a preferred embodiment of the present invention, a single digital camera (14) is used, and the projection position (4) of the slit light is photographed while changing the position of the camera.
With this method, it is possible to accurately measure the three-dimensional shape of a huge target object with a single digital camera (14).
[0013]
According to another preferred embodiment, a plurality of digital cameras (14) are used to photograph the slit light projection position (4) simultaneously from different positions. Further, the irradiation of the slit light (2, 3) and the photographing by the camera are synchronized.
This method can greatly reduce the measurement time.
[0014]
In addition, the projection position (4) of the slit light is extracted from the output of the digital camera (14), the noise component is removed, the region is thinned, and a weighted average process is performed in the perpendicular direction of the thin line to obtain the true center axis position Is calculated, and then a least-squares line for the point group is calculated.
By this method, the projection position (4) of the slit light can be obtained with high accuracy.
[0015]
  Also,The reference scale is a self-luminous type composed of a point light source and a diffusion plate located in front of the point light source.
[0016]
The slit light irradiation device (12) includes a semiconductor laser device that generates fan-shaped slit light and a scanner mirror that sequentially reflects the fan-shaped slit light to different positions.
With this configuration, it is possible to speed up the change of the projection position (4) of the slit light and to easily set a large number of projection positions (4).
[0017]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, preferred embodiments of the present invention will be described with reference to the drawings. In each figure, common portions are denoted by the same reference numerals, and redundant description is omitted.
[0018]
1. Overview
1.1 Background to development
There is a demand for a system that can easily capture the shape of parts of large structures such as ships and bridges on site. Therefore, a three-dimensional measurement technology that does not require sticking of a target was developed with the goal of conducting a measurement test on an actual structure (curved board) and evaluating its performance. By establishing this technology, it is possible to easily measure the three-dimensional shape of parts of a large structure, and it is possible to reduce assembly loss due to the occurrence of defective parts.
[0019]
1.2 Method
In the target projection type three-dimensional shape measurement of the present invention, first, a measurement object is irradiated with a lattice laser. Laser intersections and vanishing points correspond to conventional pasting targets. Subsequently, the measurement object is photographed from a plurality of viewpoints, and the positions of the laser intersections and vanishing points on the image are recognized by image processing. Finally, a three-dimensional model of the object is created by processing the image. FIG. 1 shows a configuration of a target projection type three-dimensional shape measuring apparatus of the present invention.
Advantages of the three-dimensional shape measurement of the present invention include that the target position can be easily changed, and that the number of targets can be increased or decreased. Moreover, since it is not necessary to contact an object, measurement can be performed in a short time. In addition, when measuring the shape of a part that cannot be reached, it is not necessary to assemble a scaffold for attaching the target, and the working time can be reduced.
[0020]
1.3 Apparatus
The target projection type three-dimensional shape measuring apparatus 10 of the present invention includes the following devices.
(1) Slit light irradiation device 12: a laser projector that projects lattice-shaped laser light onto an object.
(2) Digital camera 14: 6 million pixel digital camera as an image input device.
(3) Arithmetic device 16: An A4 notebook PC (CPU: 500 MHz) was adopted as the image processing device.
(4) Reference scale 18: A self-luminous scale serving as a reference for three-dimensional measurement.
[0021]
1.4 Measurement test
A bent board for hulls (about 3m x 2.5m) was used as the object 1, and a pole was installed vertically at a distance of about 5m from the bent board 1, and a camera was attached to this pole. The camera was moved on the pole, and the pole was moved to shoot from all 20 camera positions.
[0022]
1.5 Measurement test results
In order to evaluate the accuracy of the target projection type three-dimensional measurement of the present invention, a comparison was made with the measurement result by the conventional method (target sticking type).
As a result, it was concluded that the measurement accuracy of the three-dimensional position of the laser intersection is within an error of ± 1.0 mm and the standard deviation of the error is about 0.5 mm, which is almost the same as the conventional method. The required accuracy is about ± 1.0 mm, which seems to be sufficient. Moreover, sufficient accuracy could be obtained even if the number of shooting locations was reduced from 20 to about 6.
[0023]
1.6 Conclusion
The measurement accuracy of the target projection type three-dimensional measurement of the present invention is comparable to the conventional method with respect to the intersection of the laser lines.
The measurement time was about 2 hours for image capture and about 2 hours for processing. By automating the processing at one camera position, the processing time was reduced to about 50 minutes even with the current single camera system. It becomes possible to do.
In addition, by using a multiple camera system that uses a plurality of cameras at the same time, the processing time of the target projection type three-dimensional measurement can be greatly shortened. If six cameras are used, the processing time is about 5 minutes.
[0024]
2. Operation procedure
2.1 3D measurement by single camera method
In the first method of the present invention, the object on which the target is projected is photographed using one camera while changing the position of the camera, and the result is finally collected to calculate the three-dimensional position of the target. To do. Hereinafter, this method is referred to as a single camera method. FIG. 2 shows a flowchart of a three-dimensional measurement process using a single camera method. Hereinafter, the processing flow of FIG. 2 will be described.
[0025]
When the installation of the laser projector and the self-emitting scale is completed, the operator installs the camera at an appropriate position (a place where the self-emitting scale and a plurality of targets can be photographed). The operator fixes the camera at that position, projects only one of the laser lines scheduled to be projected in a grid, and shoots the situation. Subsequently, the one laser line projected earlier is erased, another laser line is projected, and the state is photographed. This is repeated for the number of laser lines.
When images of the self-emission scale and all the laser lines are taken at a certain camera position, the two-dimensional position on the screen of the target (intersection of laser lines, vanishing point, self-emission scale) is calculated from these images.
As an example, FIG. 3 shows an image for calculating the position of the intersection of two laser lines. Image processing is applied to the laser line 1 shown in the image 1 and the laser line 2 shown in the image 2, and the two-dimensional position on the image of the intersection of the laser lines 1 and 2 is calculated.
[0026]
Since the two-dimensional position of the intersection needs to be calculated with sub-pixel accuracy, it cannot be extracted by simple image processing. For this reason, the line segment approximation of each laser line is performed in the vicinity of the intersection, and the intersection of the linear equations is calculated to detect the position with sub-pixel accuracy.
Through the processing so far, the two-dimensional position of the target at a certain camera position is calculated. Subsequently, the camera position is changed, and the same shooting and processing are performed. This is repeated for the number of scheduled camera positions.
Thereby, the two-dimensional position of the target from a plurality of viewpoints can be extracted. Finally, the three-dimensional position of the target is calculated from the plurality of two-dimensional position data.
[0027]
2.2 Three-dimensional measurement by the double camera method
Next, the processing flow of the multi-camera system, which is the second method of the present invention, will be described.
The multi-camera method is a method in which a plurality of cameras are set in advance, and photographing and processing are performed simultaneously. FIG. 4 shows a state in which parallel processing is performed by four cameras. A portion surrounded by a square in the figure is a process in charge of one camera.
As described above, the processing content of the double camera method is the same as that of the single camera method, but the measurement time can be greatly shortened.
[0028]
3. Hardware specifications
3.1 Laser projector (slit light irradiation device 12)
We have newly developed a laser projector (slit light irradiation device 12) that projects sheet light emitted from an internal LD laser in a target direction through a galvanometer scanner mirror. The specifications of this device are shown below.
(1) Laser unit
・ Laser: Semiconductor laser (635 nm, less than 5 mW)
・ Class: Class 3A
・ Irradiated light: Fan shape (5.5m ahead, length 6m, width 2mm)
(2) Scanner section
・ Method: Galvanometer method with temperature control
・ Scanning axis: Independent scanning XY axis
・ Scanning angle: ± 30 °
By using this slit light irradiation device 12, it is possible to increase the speed of laser switching and increase / decrease the number of targets. The upper limit of the number of laser line intersections that can be set using this apparatus is (32768 × 2) 2 = 4.295 × 109 points. Laser light switching and galvano direction control can be controlled from the software side of the computer connected to the laser projector and RS232C.
[0029]
3.2 Self-luminous reference scale 18
Conventionally, a scale called an orientation bar has been used as a dimensional reference when calculating a three-dimensional position from a two-dimensional position on a target image. Also, a paste of reflective targets on graph paper was used as a substitute for Orient Bar. However, in these types, there is a possibility that the dimensions are distorted by heat or wind. In addition, in order to photograph the target of this orientation bar, it is necessary to irradiate flash light, and there is a problem that it cannot be integrated with the photographing of the laser line.
Therefore, a self-luminous reference scale 18 (hereinafter referred to as a self-luminous scale) has been developed so that the target position can be accurately obtained without irradiating flash light.
The specifications of this device are shown below.
(1) Light source specifications
・ Lamp: Small krypton lamp
・ Luminescent color: Visible light range
・ Optical system: Resin diffuser on the front
(2) Overall specifications
・ Cross shape size: W400 × D58 × H550mm
・ Linear size: W1000 × D58 × H50mm
[0030]
3.3 Image input device
The single-lens reflex digital camera DCS760 of Kodak was selected as the image input device. The specifications of DCS760 are as follows.
(1) Image sensor: 27.6 × 18.4 mm CCD, 6 million pixels
(2) Image quality mode: Lossless compression TIFF, non-compression TIFF
(3) Interface: IEEE1394, PCMCIAType II × 2
By using a commercially available digital camera, the system can be created at low cost. In addition, the interface has IEEE1394, and high-speed image capture is possible.
[0031]
4). Software specifications
4.1 Target 2D position calculation flow
In the method of the present invention, an image in which only the self-emission scale 18 emits light and an image in which only one of the laser lines (slit lights 2 and 3) is projected are taken from a certain camera position. By combining these images, the position of the target (self-emission scale 18, laser line intersection 5, vanishing point 6) on the image is calculated with subpixel accuracy.
In the created program, 5 vertical laser lines and 5 horizontal laser lines were projected onto the object, and the captured image was processed.
[0032]
The processing flow is shown in FIG. Hereinafter, the contents of each item in the processing of FIG. 5 will be described.
(1) Setting image processing parameters
In calculating the position of the target, it is necessary to binarize the image with a certain threshold value or to remove a region having a certain area or less. These values can be set by the user as parameters.
(2) Set image file path
Sets the path where 12 image files are stored.
(3) Creation of reduced image
The target image in this system is a 6 million pixel uncompressed TIFF file, and applying image processing as it is has many problems in terms of processing time and memory. In view of this, image processing is performed after the image size is once reduced to 1/4 in the vertical and horizontal directions (data size 1/16).
[0033]
(4) Rough self-emission scale position calculation
Binarization and noise removal are added to the reduced image, and a self-luminous scale region is extracted. Further, a weighted average process is applied to the extracted area to calculate the center position for each area. This position is added to the list as a rough position of the self-luminous target.
(5) Rough laser line vanishing point position calculation
Binarization, noise removal, and thinning processing are performed on the reduced image to extract the end points of the thin lines (points having only one effective pixel in the vicinity of the surrounding 8). The position of this point is added to the list as a rough position of the vanishing point of the laser line.
(6) Rough laser line intersection position calculation
Binarization and noise removal are added to the reduced image, and two intersection images (for example, a horizontal line reduction image and a vertical line reduction image) are ANDed to extract an intersection area. A weighted average process is added to the extracted area to calculate the center position of the area. This position is added to the list as a rough position of the intersection of the laser lines.
[0034]
(7) Detailed self-luminous scale position calculation
A part of the scale image (for example, 32 × 32 Pixel) is cut out around the position of the rough self-luminous scale. Within this area, noise components are removed and a weighted average is added to calculate the detailed position of the self-luminous scale.
(8) Detailed calculation of vanishing point of laser line
A part (32 × 32 Pixel in this case) of the 6 million pixel TIFF image showing the laser line is cut out with the position of the rough laser line vanishing point as the center. In this area, the noise component is removed and the end of the region is extracted, thereby calculating the detailed position of the vanishing point of the laser line. However, unlike the self-luminous scale and the intersection point position, the calculation accuracy is limited to Pixel accuracy.
(9) Detailed intersection calculation of laser lines
A part (32 × 32 Pixel in this case) of the 6 million pixel TIFF image in which the laser line is reflected is cut out around the position of the rough laser line intersection. Consider a case where the intersection of a horizontal line image and a vertical line image is calculated. First, a 32 × 32 pixel is cut out from the horizontal line image. Within this area, noise components are removed and the area is thinned. Further, a weighted average process is performed in the perpendicular direction of the thin line to calculate a point group that captures the true center axis position. Subsequently, a least square line for the point group is calculated. A similar process is applied to the vertical line image to calculate a least square line. Let the intersection of these two least square lines be the intersection of the laser lines.
[0035]
4.2 Laser projector control software
The laser projector (slit light irradiation device 12) used in this system can be operated by sending a command from the computer via the COM port. For the verification test, we developed control software with a simple interface. Its functions are shown below.
(1) Laser ON / OFF: The laser line can be projected / turned off by clicking a check box in the dialog.
(2) Angle change of galvanometer: By clicking a radio button in the dialog, the galvanometer can be shaken to a preset angle, and thereby the position to project the laser line can be changed.
(3) Continuous operation of the galvanometer: By pressing a button in the dialog, the galvanometer can be continuously swung with the maximum amplitude.
The interface used for the verification test is designed so that the operator can manually change the laser position, but the process of automatically changing the laser position is constructed in synchronization with camera control and image processing. It is also possible.
[0036]
5). Measurement test
5.1 Test environment
An object necessary for three-dimensional measurement is collected by placing an object on the wall of the experimental site, projecting a laser, and photographing the situation with a digital camera. It is necessary to prevent external light from entering during the shooting test. Therefore, the photographing work is performed after the surroundings are sufficiently dark.
[0037]
5.2 Data collection
Take a picture with the following procedure and collect the image data.
(1) Shooting
The camera is set at a certain position, the laser is projected 10 times (5 times vertical x 5 times horizontal), and each of them is photographed one by one for a total of 10 images. Furthermore, a single image is emitted by emitting light from the self-luminous target without projecting the laser. Therefore, 11 images are collected for one camera position.
(2) Moving the camera position
The camera position is moved to the next position, and the above (1) is performed again. This is repeated for all camera positions.
The state of target projection is shown in FIG. This figure is a composite of 10 images projected with a laser line, 1 image emitting a self-luminous target, and 1 background image taken under illumination. Further, the brightness of the laser light is emphasized by software.
The camera positions were 20 locations shown in FIG.
The number of shots is 20 (locations) × 11 (sheets) = 220 when shooting at all camera positions.
Assuming that the time required for shooting one image is 30 seconds, 220 × 30 = 6600 (seconds) = about 2 hours for shooting at all camera positions.
Since the amount of image data is about 6 MB when shooting at all camera positions, 6 (MB) × 220 (sheets) = 1320, which is about 1.3 GB.
Moreover, the measurement by the conventional method (target sticking type | formula three-dimensional measurement) was also implemented for the same bent board. A target was attached to the intersection position and vanishing point position of the laser lattice, and the three-dimensional position was measured. In the conventional method, photographing was performed from 60 camera positions. Further, as a comparative control of the accuracy of the target projection type three-dimensional measurement, images were taken from 20 camera positions by a conventional method.
[0038]
As described above, in the measurement test, measurement was performed with the following three patterns.
(1) Results of the conventional method (target sticking type) based on images taken from 60 camera positions
(2) Results of the conventional method (target sticking type) based on images taken from 20 camera positions
(3) Results of target projection type 3D measurement based on images taken from 60 camera positions
Hereinafter, the above (1) is the true value of the spatial coordinates of the target. The errors of (2) and (3) were compared and examined with respect to the true value of the spatial coordinates.
[0039]
6). Development results
6.1 Accuracy verification
(1) Intersection target accuracy
First, the accuracy of calculating the three-dimensional position of an intersection of laser lines (hereinafter referred to as an intersection target) in this method was verified.
The true value of the spatial coordinates is the three-dimensional coordinates of the intersection target calculated based on the image collected from the 60 directions around the bent plate by the conventional method. With this coordinate as positive, the accuracy of the conventional method and the target projection type will be examined.
Comparing the 20-direction image analysis error of the conventional method and the 20-direction image analysis error of the target projection type, it can be seen that the measurement accuracy is almost the same. Further, the error amount with respect to the true value of the spatial coordinates is approximately within ± 1.0 mm, the standard deviation is approximately 0.5 mm, and it can be said that the required accuracy of ± 1.0 mm is satisfied.
Moreover, the observation direction was reduced to six directions, the three-dimensional coordinates of the intersection target were calculated, and the error from the true value was calculated. Looking at the 6-direction image analysis error, the accuracy is not degraded as expected, the error amount is within ± 1.0 mm with respect to the true value of the spatial coordinates, and the standard deviation is about 0.5 mm. It can be said that
[0040]
(2) Vanishing point target accuracy
Subsequently, the accuracy of calculating the three-dimensional coordinates of the vanishing point of the laser line (hereinafter, vanishing point target) was examined.
The measurement accuracy of the vanishing point target according to the conventional method is almost equal to the measurement accuracy of the intersection target. This is because the intersection point and the vanishing point can be recognized as the same target in the conventional method. On the other hand, the detection accuracy of the vanishing point target in the target projection type was not sufficiently accurate, with a minimum error value of −8.14 mm, a maximum value of 2.71 mm, and a standard deviation of about 1.5 mm. This is because the vanishing point itself may shift due to the thickness of the camera depending on the position of the camera, and the algorithm for extracting the vanishing point position from the image with subpixel accuracy has not been established, and the extraction result is truly a vanishing point. This is thought to be because it sometimes does not capture
At present, the measurement accuracy of the vanishing point of the laser line is insufficient, but the accuracy can be improved by narrowing down the measurement object and determining in advance where the vanishing point will appear on the screen. It is possible to plan.
[0041]
(3) Intersection target accuracy in dividing and integrating observation ranges
The size of the bent plate that is the subject of this test is about 3.0 m × 2.5 m, which is enough to fit in the angle of view of the camera. However, assuming that it will be operated inline in the future, it is conceivable to perform three-dimensional measurement of a bent plate having a length of about 20 m. In such a case, the entire three-dimensional shape is calculated by shifting the imaging region while keeping the self-luminous scale serving as a reference for three-dimensional measurement and finally integrating the three-dimensional positions. . FIG. 8 shows an image of divided shooting.
[0042]
When such observation range was divided and integrated, a simple test was performed to see what the 3D position accuracy would be. From this test, the error amount with respect to the true value of the spatial coordinates is within ± 1.0 mm, and the standard deviation is about 0.5 mm, which satisfies the required accuracy. For this reason, even for a huge structure that cannot fit within the angle of view of the camera, it is possible to calculate a three-dimensional shape by creating an overlapping portion and placing a measurement reference scale there.
[0043]
6.2 Operation plan
FIG. 9 shows an assumed operation concept diagram. The effectiveness of this system is considered to be an important issue to consider how the irradiation direction can be compared with the design data of the ship. In FIG. 9, the laser sheet light in the X direction is an image of irradiating a horizontal plane in a downward direction at equal intervals in consideration of comparison with ship design data (ring cutting direction).
[0044]
6.3 Summary
The measurement accuracy of this method is comparable to the conventional method with respect to the intersection of laser lines.
Although the measurement accuracy of the vanishing point of the laser line is insufficient at present, the accuracy is improved by narrowing down the measurement target and determining in advance where the vanishing point will appear on the screen. It is possible.
The measurement time was about 2 hours for image capture and about 2 hours for processing. By automating the processing at one camera position, the processing time was reduced to about 50 minutes even with the current single camera system. It seems possible to do.
In addition, by using a multiple camera system that uses a plurality of cameras at the same time, the processing time of the target projection type three-dimensional measurement can be greatly shortened. Furthermore, as a result of the test, sufficient accuracy was obtained even with six cameras. If six cameras are used, the processing time is reduced to about 5 minutes. As a future plan, it is possible to incorporate multiple cameras and measure the three-dimensional shape of the bent plate in-line.
[0045]
In addition, this invention is not limited to embodiment mentioned above, Of course, it can change variously in the range which does not deviate from the summary of this invention.
[0046]
【The invention's effect】
As described above, the three-dimensional shape measurement method and apparatus of the present invention have excellent effects such as being able to accurately measure the three-dimensional shape of a huge target object of several m square or more in a short time. .
[Brief description of the drawings]
FIG. 1 is a configuration diagram of a target projection type three-dimensional shape measuring apparatus of the present invention.
FIG. 2 is a flowchart of a three-dimensional measurement process using a single camera method.
FIG. 3 is an image diagram for calculating a position of an intersection of two laser lines.
FIG. 4 is a flowchart of a three-dimensional measurement process using a multiple camera method.
FIG. 5 is a flowchart for calculating a two-dimensional position of a target.
FIG. 6 is a diagram showing a state of target projection.
FIG. 7 is a diagram illustrating a camera position.
FIG. 8 is an image diagram of divided shooting.
FIG. 9 is an assumed operation concept diagram.
FIG. 10 is a configuration diagram of a conventional three-dimensional position measurement apparatus.
FIG. 11 is a configuration diagram of another conventional three-dimensional position measurement apparatus.
FIG. 12 is a configuration diagram of another conventional three-dimensional position measurement apparatus.
[Explanation of symbols]
1 object, 2, 3 slit light,
4 Projection position of slit light,
5 intersections, 6 vanishing points,
10 target projection type 3D shape measuring device,
12 slit light irradiation device,
14 Digital camera,
16 arithmetic unit,
18 reference scale,

Claims (8)

Two vertical and horizontal slit lights are irradiated onto the object separately while changing the position sequentially, and the projection position of each slit light is photographed in one unit from different positions , and a reference scale serving as a dimensional reference is attached to the object. And imaging the reference scale,
Based on each image of the captured slit light, calculate the two-dimensional position of each intersection, vanishing point and reference scale of the vertical and horizontal slit light for each of the different shooting positions,
A target projection type three-dimensional shape measurement method, comprising: measuring a three-dimensional position of an object from a plurality of obtained two-dimensional position data .   The target projection type three-dimensional shape measurement method according to claim 1, wherein a single digital camera is used to photograph the projected position of the slit light while changing the position of the camera.   The target projection type three-dimensional shape measurement method according to claim 1, wherein a plurality of digital cameras are used to photograph the projection position of the slit light simultaneously from different positions.   The target projection type three-dimensional shape measuring method according to claim 1, wherein the irradiation of the slit light and the photographing by the camera are synchronized.   The projection position of slit light is extracted from the output of the digital camera, the noise component is removed, the area is thinned, weighted averaging is performed in the perpendicular direction of the thin line, and a point cloud that captures the true central axis position is calculated. The target projection type three-dimensional shape measuring method according to claim 1, wherein a least square line for the point group is calculated. The reference scale is a self-luminous type composed of a diffusion plate positioned on the front and the point light source, the target projected three-dimensional shape measuring method according to claim 1, characterized in that. A slit light irradiating device that irradiates the object separately with two vertical and horizontal slit lights while sequentially changing the position;
One or a plurality of digital cameras that shoot each slit light projection position in a single unit from different positions;
A reference scale that is affixed to the object and serves as a dimensional reference,
Based on each image of slit light photographed by the digital camera, the two-dimensional positions of the intersections, vanishing points and reference scales of the vertical and horizontal slit lights are calculated for each of the different photographing positions, and a plurality of two-dimensional positions obtained A target projection type three-dimensional shape measuring apparatus comprising: an arithmetic unit that calculates a three-dimensional position of an object from data .   The target projection type according to claim 7, wherein the slit light irradiation device includes a semiconductor laser device that generates a fan-shaped slit light and a scanner mirror that sequentially reflects the fan-shaped slit light to different positions. 3D shape measuring device.

Images