JPH0820232B2 - Three-dimensional measuring device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Outline] A three-dimensional measuring device for projecting multi-slit light to perform three-dimensional measurement. A coded multi-slit light is provided for the purpose of speeding up the three-dimensional measurement of an object to be measured. A multi-slit projector for projecting a pattern, an imaging device for imaging the coded multi-slit light pattern projected on the object to be measured, a binarization circuit for binarizing an image signal from the imaging device, and the code Each time the digitized multi-slit light pattern is switched, the weighting of the binarized image signal converted by the binarization circuit is changed, and the last weighted binarized image signal or the image signal of the previous addition result from the image memory is changed. And an image calculation unit that adds the weighted binary image signal of this time,
Based on the contour point memory for storing the coordinates of the contour point of the object to be measured corresponding to the multi-slit light decoded by the final calculation result of the image calculation unit, and the coordinates of the contour point stored in the contour point memory And a distance calculator that calculates the three-dimensional position of the contour point of the object to be measured.

[Industrial applications]

The present invention relates to a three-dimensional measuring device that projects multi-slit light to perform three-dimensional measurement.

In an automatic device such as a robot or a shape input device for inputting a three-dimensional shape, a multi-slit light is projected on an object to be measured, an image is taken by an imaging device, and a reference slit light is determined in the multi-slit light. A configuration is known in which the distance from the observation point to the contour point of the measured object is calculated. High-speed processing is demanded in such three-dimensional measurement.

[Conventional technology]

As a multi-slit projector for a three-dimensional measuring apparatus, for example, the configuration shown in FIG. 9 has been proposed previously.
That is, the laser light of a single wavelength from the semiconductor laser 51 is converted into parallel light by the collimator lens 52 and is incident on the first diffraction grating 54, and the output light 58 in which the spot lights are arranged in one row in the y-axis direction 58 And enters the second diffraction grating 55.
Since the second diffraction grating 55 is configured so that the diffraction direction is orthogonal to that of the first diffraction grating 54, the spot light becomes output light 59 arranged in a plurality of rows and is incident on the cylindrical lens 53.

The first and second diffraction gratings 54 and 55 are, for example, 20 to 70 μm.
An optical fiber array having a diameter of about m can be used.

Further, since the cylindrical lens 53 is extended in the x-axis direction, the output light 60 becomes multi-slit light in which spot lights are continuous in the y-axis direction. In this case, when the cylindrical lens 53 is extended in the y-axis direction, the output light 60 becomes multi-slit light that is continuous in the x-axis direction. This output light 60
Is incident on the shutter array 57, and the coded multi-slit light pattern 56 is controlled by opening / closing control of the selected shutter.
Becomes The shutter array 57 can be configured by, for example, a liquid crystal shutter that uses polarized light, or a shutter that uses an electro-optical effect element or the like.

FIG. 10 is an explanatory diagram of a coded multi-slit light pattern, showing a case in which different patterns A, B, and C are projected in order to distinguish eight slit lights, and A pattern is every other line and B is a pattern. The pattern is a slit light pattern with every two lines and the C pattern with every four lines. Such 3
Each time a kind of coded multi-slit light pattern is projected, the captured image signal is stored in the image memory,
For example, at the position of the slit light corresponding to the stored image signal, if it is “1” in the case of A pattern, “0” in the case of B pattern, and “1” in the case of C pattern, , B,
Since A "=" 101 ", it can be recognized that it is the slit light of number 5. That is, ⁿ coded multi-slit light patterns are projected to 2 ⁿ multi-slit light. This makes it possible to recognize all slit light numbers.

Since it is possible to identify each number of the multi-slit light projected on the DUT in this way, the three-dimensional position of each point of the DUT is calculated based on the image signal of the image of the DUT. , Three-dimensional shape can be recognized.

[Problems to be Solved by the Invention]

By projecting a coded multi-slit light pattern, the slit light can be numbered to measure the three-dimensional shape of the DUT, but as described above, in the conventional example,
The image signals corresponding to the coded multi-slit light patterns are stored in the image memories respectively, and the coded multi-slit lights are decoded, that is, the slit lights are numbered by collating the image memories. However, there is a drawback that it requires a large number of image memories.

Further, the collation processing of the image memory corresponding to the coded multi-slit light pattern is sequentially performed in units of slit light, and therefore, when projecting multi-slit light with a large number, there is a drawback that the processing time becomes long. .

An object of the present invention is to speed up three-dimensional measurement of an object to be measured.

[Means for solving the problem]

The three-dimensional measuring apparatus of the present invention is intended for high-speed processing by applying image processing, and will be described with reference to FIG.

A multi-slit projector 1 that projects a coded multi-slit light pattern by controlling a shutter array and the like, and an imaging device 2 that images the coded multi-slit light pattern projected from the multi-slit projector 1 onto an object to be measured.
And a binarization circuit 3 for binarizing the image signal from the image pickup device 2, and weighting for the binarized image signal converted by the binarization circuit 3 every time the coded multi-slit light pattern is switched. An image calculation unit 5 for changing and adding the previous weighted binary image signal or the image signal of the previous addition result read from the image memory 4 and the current weighted binary image signal, and this image calculation A contour point memory 6 for storing coordinates of contour points of the object to be measured corresponding to the multi-slit light decoded by the final calculation result of the unit 5,
The contour point memory 6 is provided with a distance calculation unit 7 for calculating the three-dimensional position of the contour point of the object to be measured based on the coordinates of the contour point.

In addition, the multi-slit projector 1 is set to the x coordinate of the image pickup device 2.
The distance calculator 7 is arranged at a position parallel to the axis, that is, at a position orthogonal to the slit light of the multi-slit light.
With a coordinate of the contour point stored in the contour point memory 6 as an address, a read-only memory (ROM) capable of reading the data of the three-dimensional position.

[Action]

The multi-slit projector 1 has a semiconductor laser, first and second diffraction gratings, a cylindrical lens, and a shutter array. By controlling the shutter array,
Coded multi-slit light patterns can be projected. The image pickup device 2 picks up an image of the object to be measured projected with the multi-slit light similarly to the television camera, and the image signal thereof is binarized by the binarizing circuit 3. In addition,
It is also possible to store the image signal in the image memory 4 once and then binarize it by the binarizing circuit 3.

The image calculation unit 5 changes the weighting for the binarized image signal every time the coded multi-slit light pattern is switched, and adds the previous weighted binary image signal and the current weighted binary image signal, Alternatively, the contents of the image memory 4 storing the previous addition result and the current weighted binary image signal are added. That is, stored in the image memory 4 the weighting of the first binarized image signal coded multi-slit light pattern as 2 ^0, a weighting of the second binarized image signals as 2 ^1, the first time And the second-time binarized image signals are added and stored in the image memory 4. Then, the weighting of the third binarized image signal is set to 2 ² ,
The result of the previous addition stored in the image memory 4 is added, and the result is stored in the image memory 4.

Similarly, n kinds of coded multi-slit light patterns are sequentially projected onto the multi-slit light composed of n slit lights to weight the i-th binarized image signal.
2 ⁱ⁻¹ is to be added to the previous addition result, and when the coded multi-slit light pattern has been projected n times and the final addition result is obtained, the decoded slit light is obtained. The binarized image signal may be weighted in the binarizing circuit 3.

Corresponding to the decoded slit light, the coordinates of the contour point of the object to be measured by multi-slit light projection are stored in the contour point memory 6, and the three-dimensional position is obtained by the distance calculation unit 7 based on the coordinates of the contour point. That is, the coordinate position of the contour point is calculated by applying the triangulation method.

In addition, the multi-slit projector 1 is set to the x coordinate of the image pickup device 2.
By arranging in a position translated in parallel to the axis, each coefficient for distance calculation can be obtained in advance, and therefore, the three-dimensional position of the measured object can be read from the coordinates stored in the contour point memory 6. It is possible to use a read-on memory (ROM) to output.

〔Example〕

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

FIG. 2 is a block diagram of an embodiment of the present invention, in which 10 is an image processor, 11 is a multi-slit projector, 12 is an image pickup device, 13 is a binarization circuit, and 14 is a region M1 to a capacity of one screen, respectively. M4 image memory, 15 image calculation unit, 16 contour point memory, 17 distance calculation unit, 18 processor (CPU) for controlling each unit, 19 main memory, 20 interface unit, 21 common bus, 22 is a multi-slit light, and 23 is an object to be measured.

The multi-slit projector 11 has, for example, as shown in FIG. 8, a semiconductor laser, first and second diffraction gratings, a cylindrical lens, and a shutter array. A controlled, coded multi-slit light pattern is measured 23
Projected on. The image pickup device 12 picks up the multi-slit light 22 projected on the object to be measured 23, and the picked-up image signal is binarized by the binarization circuit 13 and added to the image calculation section 15 or the image memory 14. . Alternatively, the captured image signal is once stored in the image memory 14 and then binarized by the binarization circuit 13.

The image processor 10 includes a binarization circuit 13, an image memory 14, an image calculation unit 15, and a contour point memory 16,
The binarization circuit 13 or the image calculator 15 weights the binarized image signal, and the multi-slit projector 11
The control information for projecting the coded multi-slit light pattern is also transferred from the processor 18 to the image processor 10, and weighting control is performed based on the control information.

FIG. 3 is a flow chart of an embodiment of the present invention, which comprises steps S1 to S11, and FIG. 4 is an operation explanatory diagram of the embodiment of the present invention. In this embodiment, the case of projecting multi-slit light composed of eight slit lights is shown.
Of course, it can be applied to the case of a large number.

First, the A pattern is projected from the multi-slit projector 11 onto the object 23 to be measured, and the image signal picked up by the image pickup device 12 is stored in the memory area M1 of the image memory 14 (S1). This A pattern is a pattern consisting of every other slit light (shown by solid lines), and the dotted line shows that it is blocked by the shutter array. Further, by imaging with the imaging device 12 from an angle different from the projection angle of the multi-slit light, the slit light is imaged in a bent state or a curved state corresponding to the shape of the DUT 23.
For example, when each slit light is imaged as a straight line,
You can see that it is a plane.

Next, the A pattern image signal stored in the memory area M1 is binarized by the binarizing circuit 13, weighted and stored in the memory area M2, and the image signal IA is obtained (S2). In this case, the low level "0" is 0 and the high level "1" is
It is an ^{2 0 = 1 (H = 2} 0). Therefore, as shown in the pattern A of FIG. 4, the image signal IA is composed of the multi-slit light represented by “10101010”.

Next, the B pattern is projected and the captured image signal is stored in the memory area M3 (S3). This B pattern is a pattern composed of every other slit light.

The image signal stored in the memory area M3 is binarized by the binarization circuit 13. Weighting is performed by the binarization circuit 13 or the image calculation unit 15. The high level "1" of the binarized image signal is set to 2 ¹ = 2 (H = 2 ¹ ), and the image is stored in the memory area M4. Signal IB (S4). Therefore, as shown in the pattern B of FIG. 4, the image signal IB is the multi-slit light represented by "2202200".

The image signal IA stored in the memory area M2 and the image signal IB stored in the memory area M4 are read out and added by the image calculation section 15, and the result is stored in the memory area M1 to obtain the image signal IS. (S5). Therefore, as shown in FIG. 4, the image signal IS becomes the image signal IS composed of the multi-slit light represented by "32103210".

Next, the C pattern is projected and the captured image signal is stored in the memory area M3 (S6). This C pattern is a pattern composed of every four slit lights.

The image signal stored in the memory area M3 is binarized by the binarization circuit 13, and 2 ² = 4 is weighted for the high level “1” of the binarized image signal (H = 2 ² ). Store in area M4 and use it as image signal IC (S
7). Therefore, as shown in the pattern C in FIG.
Image signal IC consisting of multi-slit light represented by 40000 "
Becomes

Next, the image signal IS stored in the memory area M1 and the image signal IC stored in the memory area M4 are compared with each other by the image calculation unit.
Add by 15, store the result in memory area M2,
Image signal IS '(S8). Therefore, as shown in FIG. 4, the image signal of the addition result becomes the image signal IS 'consisting of the multi-slit light represented by "76543210", and the coded multi-slit light is decoded. The encoding in this case shows the case of natural binary numbers, but it is also possible to use other codes such as Gray code.

Since the image signal IS 'stored in the memory area M2 showing the decoding result is a multi-valued number, it is binarized,
The coordinate points of 2 ⁰ or more levels stored in the contour point memory 16 (S9). Then, the contents of the contour point memory 16 are read (S10), and the distance calculation unit 17 calculates the distance for each coordinate point (S11). This distance calculation is performed by the dedicated distance calculation unit 1
Although the case is shown by 7, the calculation function of the processor 18 can also be used. Further, when various parameters and the like are determined, the distance calculation unit 17 can be configured by a read only memory (ROM).

The outline of the distance measurement will be described with reference to FIG. The camera coordinate system whose origin is the lens center of the imaging device 12 is O-XY.
Letting Z be the light source coordinate system with the light source center of the multi-slit projector 11 as the origin, the relationship between the two is It is represented by. Here, t _ij (i = 1 to 3, j = 1 to 4) is
It is a coefficient determined by the positional relationship between the multi-slit projector 11 and the imaging device 12, and can be calculated.

The multi-slit projector 11 defines each slit light surface composed of m slit lights spread around the y-axis by π _j (j = 1 to 1).
m). Focusing on the j-th slit light, the projection image P is formed on the object to be measured by the projection, and the picked-up image I is formed on the image plane π ₁ of the image pickup device 12. At that time, the three-dimensional position of the point P _k (X _k , Y _k , Z _k ) on the projected image is based on the principle of triangulation and the point I _k (x _k , y on the lens center O and the image plane π _{I. It} can be calculated as the intersection of the line of sight OI _k connecting _k ) and the slit light plane π _j .

u = h / g (3) g = (t ₁₁ X _k + t ₁₂ y _k + t ₁₃ f) cos θ _j − (t ₃₁ X _k + t ₃₂ y _k
+ T ₃₃ f) sin θ _j (4) h = t ₃₄ sin θ _j −t ₁₄ cos θ _j (5) where (x _k , y _k ) indicates the position of P _k on the image plane,
θ _j represents the projection angle of the slit light surface π _j , and f is the focal length.

(X _k , y _k ) is obtained as the coordinates stored in the contour point memory 16, and θ _j is the decoded slit light number j.
Can be obtained from That is, the value of j is the contour point memory
It can be calculated as the gray level stored in 16.

FIG. 6 is a block diagram of a main part of the distance calculation unit 17, in which the coordinates x _k and y _k read from the contour point memory 16 and the focal length f are input, and the slit light numbers (1) to (m) are input. By the corresponding calculation units 31-1 to 31-m, u according to the above equation (3)
(1) to u (m) are calculated. Then, the slit light number (1) to
U (j) corresponding to (m) is added to the multiplication units 33 to 35,
The multiplication according to the equation (2) is performed, and the three-dimensional position X _k , Y _k , Z _k
Is required.

In this case, by arranging the multi-slit projector 11 in a position parallel to the X-axis of the coordinates of the image pickup device 12, u in the equation (3) is expressed by u = α / (β + γ) (6) α ＝ t ₃₄ sinθ _j −t ₁₄ cosθ _j …… (7) β ＝ t ₁₁ cosθ _j ＋ t ₁₃ sinθ _j …… (8) γ ＝ t ₁₃ cosθ _j −t ₁₁ sinθ _j …… (9) it can. Then, the coefficients t _{11 to} t ₁₄ , t _{31 to} t ₃₄
Is a value that can be obtained in advance as described above, and θ _j is m according to the slit light numbers (1) to (m).
However, u is a function of x _k and θ _j , so that the coefficients α, β, and γ corresponding to θ _j can be obtained in advance. ) To u
The three-dimensional position of the object to be measured 23 can be quickly obtained by storing the value of X or the values of X _k , Y _k , and Z _k .

FIG. 7 is a diagram for explaining the operation of the pipeline processing according to the embodiment of the present invention. F1 to F8 indicate frames of the image signal, and in the frame F1, the A pattern is projected onto the object 23 to be measured, and the image pickup apparatus is used. The image signal of 12 is stored in the memory area M1.
In the next frame F2, the A pattern image signal of the memory area M1 is binarized and stored as the image signal IA in the memory area M2, and the B pattern is projected on the object to be measured 23, and the imaging device
The image signal from 12 is stored in the memory area M3.

In the next frame F3, the B pattern image signal in the memory area M3 is binarized and stored in the memory area M4 as the image signal IB.

In the next frame F4, the contents of the memory areas M2 and M4 are added and stored in the memory area M1 as the image signal IS.

In the next frame F5, the C pattern is projected onto the DUT 23, the image signal from the imaging device 12 is stored in the memory area M3, and in the next frame F6, the C pattern image signal in the memory area M3 is binary. It is stored in the memory area M4 as an image signal IC.

In the next frame F7, the contents of the memory areas M1 and M4 are added and stored in the memory area M2 as the image signal IS ′. This means that the coded multi-slit light has been decoded. Then, in the next frame F8, the image signal IS ′ in the memory area M2 is binarized, and the contour point memory
Store at 16.

The decoding process in this embodiment requires 7 frame periods in the case of 8 slit lights, but 9 frame periods in the case of a larger number of slit lights, for example, 16 slit lights. Therefore, in the case of 32 lines, it is possible to decode each in 11 frame periods. Therefore, as the number of slit light beams increases, it is possible to decode with a slight increase in the number of frames.

FIG. 8 is an operation explanatory diagram of the pipeline processing of another embodiment of the present invention, in which the pattern A is projected in the frame F1, the image signal from the image pickup device 12 is stored in the memory area M1, and the next frame is displayed. At F2, the A pattern image signal of the memory area M1 is binarized and stored as the image signal IA in the memory area M2, and at the same time, the B pattern is projected onto the object 23 to be measured, and the image signal from the image pickup device 12 is stored in the memory area M3. The processing to be stored in is similar to that of the embodiment shown in FIG. 7, but in the next frame F3, the processing in the frames F3, F4 and F5 of FIG. 7 is performed simultaneously. , C pattern to be measured 23
The image signal from the image pickup device 12 is projected onto the memory area M4.
, The B pattern image signal in the memory area M3 is binarized, the image signal IA in the memory area M2 is read out, and the addition is performed with a delay corresponding to the binarization processing.
The image signal IS is stored in.

In the next frame F4, the frames F6 and F7 shown in FIG.
The memory area M4
The C pattern image signal is binarized, the image signal IS in the memory area M1 is read out, a delay corresponding to the binarization processing is performed, addition processing is performed, and the image signal IS 'is stored in the memory area M2. By this, the coded multi-slit light is decoded. In the next frame F5, the image signal IS 'in the memory M2 is binarized and stored in the contour point memory 16, as in the frame F8 in FIG. In this frame F5, the step of projecting the pattern A for the next measurement of the object to be measured can be started.

Further, the image signals read from the memory areas M2 and M1 in the above-mentioned frames F3 and F4 give the address signals of the respective memory areas M1 to M4 in common, so that the time required for the binarization processing is delayed. However, it is also possible to delay the read address of the image signals IA and IS to perform the addition process between the binarized image signals.

In this embodiment, the binarization of the image signal, the inter-image addition processing, and the projection of the coded multi-slit light are performed simultaneously, so that in the case of eight slit light, as described above, Frame, or 5 frames for 16 slit light, 6 frame for 32 slit light,
Each can be decoded. That is, the decoding process can be performed in a shorter time than the embodiment shown in FIG.

The present invention is not limited to the above-described embodiments, but various additions and modifications can be made.

〔The invention's effect〕

As described above, according to the present invention, the coded multi-slit light is projected from the multi-slit projector 1 onto the object to be measured, and the image signal picked up by the image pickup device 2 is converted into the binarization circuit 3.
Each time the pattern of the coded multi-slit light is switched, the weighting of the binarized image signal is changed, and the last binarized image signal or the image signal of the previous addition result and the image calculation unit 5 are used. By performing the addition process, the coded multi-slit light can be decoded, and the capacity of the image memory 4 can be reduced to, for example, about 4 screens even if the number of slit lights of the multi-slit light is large. Therefore, it is possible to reduce the size and make it economical. Further, by adding between images by an image processor or the like, the coded multi-slit light can be decoded at high speed, thereby speeding up the three-dimensional measurement processing.

Further, by arranging the multi-slit projector 1 at a position which is moved in parallel on the coordinate X-axis of the image pickup device 2, it becomes possible to previously obtain the coefficient and the like in the distance calculation. Therefore, the read-on memory is used. Since three-dimensional coordinates can be read out by means of the three-dimensional coordinate system, there are advantages that the configuration is simplified and high-speed processing is possible.

[Brief description of drawings]

FIG. 1 is an explanatory view of the principle of the present invention, FIG. 2 is a block diagram of an embodiment of the present invention, FIG. 3 is a flowchart of the embodiment of the present invention, and FIG. 4 is an operation explanatory view of the embodiment of the present invention. FIG. 5 is an explanatory view of distance measurement, FIG. 6 is a block diagram of a main part of the distance calculation unit,
FIG. 7 is an operation explanatory diagram of pipeline processing of one embodiment of the present invention, FIG. 8 is an operation explanatory diagram of pipeline processing of another embodiment of the present invention, and FIG. 9 is the previously proposed multi-slit. FIG. 10 is an explanatory diagram of the projector, and FIG. 10 is an explanatory diagram of a coded multi-slit light pattern. 1 is a multi-slit projector, 2 is an image pickup device, 3 is a binarization circuit, 4 is an image memory, 5 is an image calculation unit, 6 is a contour point memory, and 7 is a distance calculation unit.

Claims (2)

[Claims] 1. A multi-slit projector (1) for projecting a coded multi-slit light pattern, an imaging device (2) for imaging the coded multi-slit light pattern projected on an object to be measured, and the imaging device ( A binarization circuit (3) for binarizing the image signal from 2), and a binarized image signal converted by the binarization circuit (3) every time the coded multi-slit light pattern is switched. An image calculation unit (5) for changing the weighting and adding the previous weighted binary image signal or the image signal of the previous addition result read from the image memory (4) and the current weighted binary image signal A contour point memory (6) for storing the coordinates of the contour point of the measured object corresponding to the multi-slit light decoded by the final calculation result of the image calculation section (5); Based on the coordinates of the stored contour point Li (6), the three-dimensional measuring apparatus characterized by comprising distance calculation unit (7) for calculating the three-dimensional position of the contour points of the object. 2. The multi-slit projector (1) is arranged at a position translated in parallel on the x-axis of the coordinates of the image pickup device 2, and the distance calculator (7) is stored in the contour point memory (6). The three-dimensional measuring apparatus according to claim 1, wherein the three-dimensional measuring apparatus is configured by a read-only memory that reads out a three-dimensional position based on the coordinates of the contour points that have been created.