JP2003315028A - Three-dimensional measuring apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To automatically discriminate a loss in three-dimensional data and to automatically recover the data in a three-dimensional measuring apparatus for measuring an object shape without any contact by a light cutting method using slit beams. <P>SOLUTION: The three-dimensional measuring apparatus comprises: a means for converting image data composed of slit-like reflection light; and a discriminating means for discriminating the continuity of image data that are converted to vector. In a spare scan (S1), the continuity of vector obtained by converting slit images to vector for each scan is examined (S2). When the slit images are not continuous (NO for S2), a rotary stand where the subject is placed is rotated until the slit images become continuous (S5, S6, S7). Then, in a main scan (S9), the three-dimensional data are measured, a scan angle where the three-dimensional data become defective is stored (S12), and the rotary stand is rotated at the angle (S16), thus automatically recovering the three-dimensional data. <P>COPYRIGHT: (C)2004,JPO

Description

Description: BACKGROUND OF THE INVENTION [0001] 1. Field of the Invention
Non-contact measurement of the shape of an object by irradiating light such as 3rd order
For the former measuring device. 2. Description of the Related Art Conventionally, light cutting for performing slit light projection has been performed.
Non-contact type three-dimensional measuring device by the method
You. Non-contact type 3D measuring device is faster than contact type
CG system and CAD system
Data input to stem, body measurement, visual recognition of robot
It is used for such purposes. [0003] As a measurement method suitable for this three-dimensional measurement, it is known.
Slit light projection method (also called light cutting method)
3D by optically scanning an object (subject)
An image is a method of obtaining a so-called distance image, and a specific detection light
Is an active measurement method that shoots an object by irradiating
You. Then, the object is placed on the rotary table on the back side,
It is fixed to the near side while rotating the object appropriately
The shape of the object is measured by the imaging means. 3D image
Is a set of pixels indicating the three-dimensional positions of multiple parts on the object
It is. In the slit light projection method, the cross section is straight as the detection light.
Linear slit light is used. FIG. 21 shows an outline of the slit light projection method.
FIG. 22 shows the principle of measurement by the slit light projection method. Total
A band-shaped slit light U having a narrow cross section is applied to the object Q to be measured.
Irradiate and reflect the reflected light with a two-dimensional image sensor, for example.
The light is incident on the image plane S2 (FIG. 21A). Irradiation of object Q
If the part is flat, the slit image that is the shot image is a straight line
[FIG. 21 (b)]. If the irradiated area has irregularities,
The direction in which the lit light U is irradiated differs from the direction in which the reflected light is received.
Therefore, the slit image reflects the unevenness of the irradiated part,
An image in which a straight line bends or becomes step-like [Fig.
1 (c)]. That is, the distance between the measuring device and the object Q is large or small.
Reflects the incident position of the reflected light on the imaging surface S2 [FIG.
21 (d)]. The slit light U is deflected in the width direction.
Scans the surface of the object within the range visible from the light receiving side
Can sample the three-dimensional position of the object surface
You. The number of sampling points depends on the number of pixels of the image sensor
I do. In FIG. 22, a starting point A of light emission and a light receiving system
Base line AO connecting the principal point O of the lens is perpendicular to the light receiving axis
Thus, a light projecting system and a light receiving system are arranged. The receiving axis is
It is perpendicular to the imaging surface S2. The main points of the lens
Is when the image of the finite object is formed on the imaging surface S2
Of the imaging surface S by the so-called image distance (b)
It is a point on the light receiving axis away from 2. The image distance b is a light receiving system
Between the focal length f of the lens and the amount of lens extension for focus adjustment
It is sum. The principal point O is defined as the origin of a three-dimensional rectangular coordinate system.
The light receiving axis is the Z axis, the base line AO is the Y axis, and the length direction of the slit light
Is the X axis. When the slit light U reaches a point P (X, Y,
Z) and the light projection axis and the light emission reference plane (irradiation axis
Θa and the light receiving angle θp,
The coordinate Z of the point P is represented by the equation (1). Baseline length L = L1 + L2 = Z × tan θa + Z × tan θp∴Z = L / (tan θa + tan θp) (1) The light receiving angle θp is a straight line connecting the point P and the principal point O.
This is an angle formed with a plane including the light receiving axis (light receiving axis plane). Since the imaging magnification β is β = b / Z,
The distance in the X direction between the center of the image plane S2 and the light receiving pixel is xp, Y
Assuming that the distance in the direction is yp (see FIG. 22A), the point P
Are represented by equations (2) and (3). X = x / β (2) Y = yp / β (3) The angle θa is univocally determined by the angular velocity of deflection of the slit light U.
Is decided. Is light-receiving angle θp related to tan θp = b / yp
Can be calculated from In other words, the position (xp,
yp), it is possible to determine the angle based on the angle θa at that time.
Accordingly, the three-dimensional position of the point P can be obtained. [0010] As shown in FIG.
In the case where a group of noises is provided, the principal point O becomes the rear principal point H ′.
Assuming that the distance between the rear principal point H ′ and the front principal point H is M, the point P
Is represented by equation (1B). L = L1 + L2 = Z × tan θa + (Z−M) × tan θp∴Z = (L + M × tan θp) / (tan θa + tan θp) (1B) In the measurement by the slit light projection method based on the above principle, an example is given.
For example, is the imaging surface S2 a finite number of pixels like a CCD sensor?
When using imaging means consisting of
It depends on the pixel pitch of the means. However, on the imaging surface S2
The width of the slit light U in the Y direction (scanning direction) is equivalent to a plurality of pixels.
By setting the slit light U so that
Performance can be enhanced. Conventionally, in such a measuring device, an object (subject)
Object on the rotary table on the back side and rotate the subject
Of the object with the imaging means fixed to this side.
The shape is being measured. As another conventional example, the resolution is two-dimensional.
It is known to display a bar below the image. During measurement
User judged dissatisfied with the indicated resolution
In the case, image data of high magnification is obtained through a long focal length lens.
This makes it possible to obtain image data with accuracy appropriate for the shape of the target object.
Generated. At this time, the turntable on which the object to be measured is
Operate to measure the field of view divided into multiple areas
(See Japanese Patent Application Laid-Open No. 7-174537). [0014] However, light cutting
If the object has a complicated shape in the method, the slit light
When it does not hit the subject or reflected light from the object
Cannot be read because it is blocked by part of the object on the optical path of the reflected light
May occur. Therefore, during one scan, the shadow area
Minute, that is, a part where 3D data is missing appears
Problem. In the past, we moved objects manually to deal with them.
And it is difficult to automatically recognize them. Light cutting method
When recognizing a more three-dimensional shape,
Blind spots (shades) are inevitable.
Technology for efficiently recognizing shapes without relying on movement is essential
Yes, its realization was desired. The present invention solves the above problems.
Therefore, when a part where three-dimensional data is lost appears during scanning,
Automatically finds that part and deletes the missing part.
A three-dimensional measurement device that can automatically repair data
To provide. [0016] To achieve the above object,
According to the first aspect of the present invention, a subject is projected by projecting slit light.
Projection means for optically scanning the
Reflection of the slit light projected on the subject from the subject
Light receiving means for receiving light, and rotation for placing the subject
And a drive mechanism for rotating the turntable.
In a three-dimensional measuring device that measures the shape of the body without contact,
Two-dimensional image composed of slit-like reflected light received by the light receiving means
Means for converting the image data into vectors, and
Determining means for determining the continuity of the image data;
The obtained image data is converted to a vector
Vector conversion by means of conversion, this vector conversion
The continuity of the obtained image data is determined by the determination means,
If it is determined that the continuity is lacking, the slit
The turntable is rotated by the drive mechanism at the light scanning position.
To rotate the subject, and again project the slit light.
This is to receive the reflected light of the shadow and the projection light. In the above configuration, the anti-slit light
Vector conversion of the two-dimensional image of the light, this vector conversion
Continuity of data, that is, three-dimensional data
Automatically determines if there is any data loss, and
Rotating the subject so that continuity can be achieved
The table is rotated by the drive mechanism, and the slit light is projected and
And reflected light of projection light, so there is no defect
Three-dimensional data can be collected automatically. FIG. 1 shows a measuring system according to the present invention.
1 is shown. The measurement system 1 uses a slit light projection method.
3D camera (Range Fine)
2) a host for processing output data of the 2 and 3D camera 2
And a turntable 5 for rotating the object Q as a subject.
It is composed of The three-dimensional camera 2 includes a plurality of suns on the object Q.
Measurement data that identifies the three-dimensional position of the pulling point (slip
2 indicating the color information of the object Q together with the image data
Outputs 3D images and data required for calibration
I do. Find coordinates of sampling points using triangulation
The arithmetic processing is performed by the host 3. The host 3 includes a CPU 3a, a display 3
b, keyboard 3c, mouse 3d, etc.
Computer system. The CPU 3a has a measurement data
Data processing software is installed. E
Between the strike 3 and the 3D camera 2
Both forms of offline with portable recording media 4
Data can be transferred between the host 3 and the turntable 5
In, data transfer in an online form is possible.
As the recording medium 4, a magneto-optical disk (MO),
There are two disks (MD) and memory cards. FIG. 2 shows the appearance of the three-dimensional camera 2. Howe
A light emitting window 20a and a light receiving window 20b are provided in front of the jing 20.
Have been killed. The light emitting window 20a is positioned above the light receiving window 20b.
Located on the side. Pick-up from the internal optical unit OU
Light (a band-shaped laser beam having a predetermined width w) U
It goes to the object (subject) to be measured through 20a. S
The radiation angle φ in the length direction M1 of the lit light U is fixed.
A part of the slit light U reflected on the surface of the object is
b and enters the optical unit OU. The optical unit
The knit OU optimizes the relative relationship between the projecting axis and the receiving axis.
A two-axis adjustment mechanism for On the upper surface of the housing 20, a zooming
Tons 25a, 25b, manual focusing button 26
a, 26b, and a shutter button 27 are provided.
You. As shown in FIG. 2B, on the back of the housing 20,
Liquid crystal display 21, cursor button 22, select
Button 23, cancel button 24, analog output terminal
31, 32, digital output terminal 33, and recording medium
Four attachment / detachment ports 30a are provided. The liquid crystal display 21 (LCD) operates
It is used as a screen display means and an electronic finder.
The photographer can set the shooting mode by using the buttons 22 to 24 on the back.
Can be set. From analog output terminal 31
Outputs the measurement data, and the analog output terminal 31
A two-dimensional image signal is output, for example, in the NTSC format. De
The digital output terminal 33 is, for example, a SCSI terminal. FIG. 3 shows a functional configuration of the three-dimensional camera 2.
Solid arrows in the figure indicate the flow of electric signals, and broken arrows indicate light.
It shows the flow. The three-dimensional camera 2 includes the optical unit described above.
Two optical systems on the light emitting side and the light receiving side constituting the knit OU
40 (projection means) and 50 (light receiving means). light
In the science system 40, a semiconductor laser (LD) 41 emits light.
Laser beam having a wavelength of 670 nm
Pass through the slit light U, and the galvano
The light is deflected by a mirror (scanning means) 43. Semiconductor
The driver 44 of the user 41 and the driving system 4 of the projection lens system 42
5 and the drive system 46 of the galvanometer mirror 43
It is controlled by the controller 61. In the optical system 50, a zoom unit 51
Is collected by the beam splitter 52.
It is split. The light in the oscillation wavelength band of the semiconductor laser 41 is
The light enters the sensor 53 for measurement. Light in the visible band
Incident on the color sensor 54 for the color filter. Sensor 53 and power
Color sensors 54 are both CCD area sensors.
You. The zoom unit 51 is a focus type, and a part of the incident light.
Are used for auto-focusing (AF). AF machine
The function is based on the AF sensor 57, the lens controller 58, and the camera.
This is realized by the focusing driving system 59. zooming
The drive system 60 is provided for electric zooming. The imaging information from the sensor 53 is transmitted to the driver 5
5 is transferred to the output processing circuit 62 in synchronization with the clock from
It is. Each pixel of the sensor 53 by the output processing circuit 62
Is generated and stored in the memories 63 and 64.
Is stored. The operator then directs the data output
Then, the measurement data is stored in the SCSI controller 66 or N
Online output in predetermined format by TSC conversion circuit 65
Or stored in the recording medium 4. Measurement data
The analog output terminal 31 or digital
A terminal output terminal 33 is used. By the color sensor 54
Image information is synchronized with the clock from the driver 56.
The data is transferred to the color processing circuit 67. Color processed
The imaging information is sent to an NTSC conversion circuit 70 and an analog output terminal.
Output on-line via digital device 32 or digital image
Quantized by the generation unit 68 and stored in the color image memory 69
Is done. After that, the color image data is stored in the color image memory.
69 to the SCSI controller 66,
Output from the output terminal 33
The data is stored in the recording medium 4 in association with the data. In addition,
The color image has the same angle of view as the distance image obtained by the sensor 53.
And application processing on the host 3 side
It will be used as reference information. Use color information
Processing, for example, a plurality of sets of different camera viewpoints
Generate 3D shape model by combining measurement data
Processing, thinning out unnecessary vertices of 3D model, etc.
There is. The system controller 61
The appropriate text is displayed on the screen of the LCD 21 with respect to the
Give instructions to display letters and symbols. FIG. 4 shows the structure of the light projecting lens system 42. Throw
The optical lens system 42 includes a collimator lens 421,
Lens 422 and expander lens 423
Lens. The semiconductor laser 41 emits
The appropriate slits in the following order for the laser beam
Optical processing for obtaining light U is performed. First, Kolime
The beam is collimated by the data lens 421. next
Beam of laser beam by variator lens 422
The diameter is adjusted. Finally, the expander lens 423
As a result, the beam is expanded in the slit length direction M1. The variator lens 422 is used to determine the photographing distance and
Regardless of the angle of view of photographing, three or more pixels
It is provided to make slit light U of minute width incident.
You. The drive system 45 receives instructions from the system controller 61.
Therefore, the width w of the slit light U on the sensor 53 is kept constant.
The variator lens 422 is moved to keep it. Bari
The eta lens 422 and the zoom unit 51 on the light receiving side
Interlock. Before the deflection by the galvanomirror 43,
Increasing the lit length allows for a better
All distortions of the slit light U can be reduced. Eki
The spanner lens 423 is arranged at the last stage of the projection lens system 42.
Position, that is, close to the galvanometer mirror 43.
To reduce the size of the galvanometer mirror 43
Can be. FIG. 5 shows a zoom unit 51 for receiving light.
The schematic configuration is shown. The zoom unit 51 includes the front image forming unit 5
15, variator section 514, compensator section 513,
Focusing unit 512, rear image forming unit 511, and incidence
Beam splitter 51 that guides a part of light to AF sensor 57
6. Front imaging unit 515 and rear imaging
The part 511 is fixed with respect to the optical axis. The movement of the focusing unit 512 is
Sing drive system 59 is responsible for moving variator unit 514.
The zooming drive system 60 is responsible. Focusing drive system 5
9 is the moving distance of the focusing unit 512
Focusing encoder 59A indicating the
ing. The zooming drive system 60 includes a variator unit 514.
Zooming point indicating the moving distance (zoom step value)
A coder 60A is provided. FIG. 6 shows a schematic configuration of the beam splitter 52.
FIG. 7 shows the light receiving wavelength of the sensor 53 for measurement.
Reference numeral 8 denotes a light receiving wavelength of the monitor color sensor 54. The beam splitter 52 is a color separation film (die)
Clock mirror) 521 and two
Prisms 522, 523, exit surface 52 of prism 522
Infrared cut filter 524 and sensor provided in 2b
53, a visible cut filter 525 provided on the front side of
An infrared camera provided on the exit surface 523b of the prism 523
Filter 526 and low-pass filters 527 and 5
28. Light UC incident from the zoom unit 51
Passes through the low-pass filter 527 and the prism 522
The light enters the color separation film 521. Oscillation band of semiconductor laser 41
The light U0 in the area is reflected by the color separation film 521, and the prism 522
After being reflected by the incident surface 522a of the
Injects toward the sensor 53. Emitted from prism 522
Out of the light U0, the infrared cut filter 524 and the visible light
The light transmitted through the cut filter 525 is detected by the sensor 53.
Received. On the other hand, the light C0 transmitted through the color separation film 521
Is color from the exit surface 523b through the prism 523
Inject toward the sensor 54. Fire from prism 523
Of the emitted light C0, the infrared cut filter 526 and the
The light transmitted through the multi-pass filter 528 is
Is received by the In FIG. 7, the color components are
The dissolving film 521 sets the wavelength of the slit light (λ: 670 nm).
Reflects light in a relatively wide range of wavelength bands. Toes
The wavelength selectivity of the color separation film 521 is such that only the slit light
It is insufficient to make the light incident on the sensor 53 selectively. And
However, in the beam splitter 52, the characteristic indicated by the chain line
Infrared cut filter 524 and the characteristics indicated by the solid line
Since the visible cut filter 525 is provided,
The light finally incident on the sensor 53 is indicated by hatching in FIG.
This is light of a narrow range of wavelengths indicated by. This makes the ring
A meter with a small influence of boundary light, that is, a large optical SN ratio
Measurement can be realized. On the other hand, the color sensor 54 has a solid line in FIG.
The infrared cut filter 528 having the characteristics indicated by
And red transmitted through the color separation film 521 having the characteristics indicated by the broken line.
Since the light in the outer band is blocked, only visible light enters.
Thereby, the color reproducibility of the monitor image is improved. Note that the infrared cut filter 524 and the
Instead of using two filters of visual cut filter 525
In addition, one piece of infrared and visible light blocking properties
A filter may be used. Infrared cut filter 524 and
And the visible cut filter 525
Or both filters may be connected to the sensor 53
May be provided on the side of. In contrast to the example of FIG.
A filter 525 is provided on the side of the prism 522, and
The filter 524 may be provided on the sensor 53 side. Next, the two axes provided in the optical unit OU
The adjustment mechanism will be described with reference to FIGS.
As shown in FIG. 9, the optical unit OU is a projection unit.
Optical system 40 and optical system 50 as a light receiving means
It is configured to be attached to the units 211 and 212.
These two brackets 211, 212 are
Are connected to each other so as to be rotatable about a certain second rotation axis AX2.
Have been. The optical system 40 moves with respect to the bracket 211.
It is rotatable about the first rotation axis AX1, which is the Z-direction axis.
Is attached. The optical system 50 is fixed to the bracket 212.
Is defined. The first rotation axis AX1 is received by the light receiving optical system 50.
It is adjusted so as to be parallel to the optical axis AX3. As shown in FIG. 10 to FIG.
The units 211 and 212 have a substantially L-shaped side view.
The outer surfaces of the horizontal plate portions 211a and 212a are
Can be rotated in a state of contact. That is, horizontal plate
The collar 216 is turned in the hole 215 provided in the portion 212a.
The collar 216 is rotatably fitted, and the collar 216 is
Is fixed to the horizontal plate portion 211a. Bolt 2
17 has a screw hole in the head,
After the bottom cylindrical cap is put on the head, the cap
Through the hole provided in the center of the head and screw it into the screw hole on the head
Bolt 217, thereby securing the bolt 217.
The head is covered. In addition, the head of the bolt 217
A groove for rotational engagement is provided. At the protruding end 218 of the horizontal plate 212a
Screw holes have adjustments to adjust the rotation angle position
The bolt 219 is screwed. Tip of adjustment bolt 219
The part is attached to the horizontal plate part 211a with the bolt 221.
Abuts on the peripheral surface of the collar 222 provided. The bolt 22
1 and a bolt 223 attached to the horizontal plate portion 212a.
Between them, a tension spring 224 is mounted.
Therefore, the adjustment is performed between the horizontal plate portions 211a and 212a.
In the direction in which the tip of the bolt 219 contacts the collar 222
Are biased to each other. Therefore, the adjustment bolt 219
To adjust its axial position,
Bracket 211 and bracket around two axes of rotation AX2
The rotation angle position relative to the slot 212 is adjusted. Key
After adjusting the adjusting bolt 219, lock the adjusting bolt 219.
While fixing with the nut 220, the horizontal plate portion 212a
The horizontal plate portion 211 penetrates the three elongated holes 225 provided.
Tighten the three bolts 226 screwed into the screw holes a
With this, the space between the two horizontal plate portions 211a and 212a is fixed.
You. A shaft section is provided on the rear side of the housing of the optical system 40.
A member 231 is attached, and this shaft member 231 is
The first rotation axis AX1 is centered on the vertical plate portion of the bracket 211.
Is rotatably fitted in a shaft hole 232 provided in
You. Rotation angle of optical system 40 about first rotation axis AX1
After adjusting the angle position, it is provided on the housing of the optical system 40.
Screw holes provided in the bracket 211 through the holes
By tightening several bolts (not shown)
Thus, the optical system 40 is fixed to the bracket 211. bra
A mounting plate 213 is fixed to the bracket 212 with bolts.
The mounting plate 213 is a casing of the optical unit OU.
Attached to. The starting point of light emission in the light emitting optical system 40
A and the principal point O of the lens of the light receiving optical system 50 (rear principal point H ')
Are perpendicular to the light receiving axis AX3. Imaging
The surface S2 is perpendicular to the refracted light receiving axis AX3. Next, the first rotation axis AX1 and the second rotation axis A
A method of adjusting X2 will be described. As shown in FIG.
The screen SCR is received in front of the light receiving axis AX3.
It is arranged perpendicular to the optical axis AX3. First, the projection optical system
Slit light U projected on screen SCR from 40
When scanning with slit light U, before and after scanning
Left and right moving distances AL1 and AL2 of slit light U at
Are adjusted so that are the same as each other.
You. Next, light is received on the imaging surface S2 shown in FIG.
Left and right positions BL1 and BL of the slit light U
2 are the same as each other, that is, the slit light U
The first rotation axis AX1 is set so as to be parallel to the X axis of the image plane S2.
To adjust. Repeat these adjustments several times. By these adjustments, the first rotation axis AX1
Is parallel to the light receiving axis AX3, and the scanning direction of the slit light U
(Deflection direction) coincides with the direction of the second rotation axis AX2. And
Therefore, the error in the positional relationship between the optical system 40 and the optical system 50
And accurate measurement can be performed without correction.
It can be carried out. Also, to obtain better accuracy,
Even when performing the correction, the zoom unit 51
It is not necessary to change the correction value even after performing the tuning. did
Therefore, there is no need or minimal computational processing for correction.
And the processing time becomes extremely short. FIG. 14 shows a three-dimensional position in the measurement system 1.
The principle of calculation of the position is shown. In the figure, it is easy to understand
Therefore, the same reference numerals are given to the elements corresponding to FIG. 21 and FIG.
It is attached. A plurality of pixels on the imaging surface S2 of the sensor 53
Irradiates the object Q with a relatively wide slit light U
You. Specifically, the width of the slit light U is set to 5 pixels. S
The lit light U is 1 on the imaging surface S2 for each sampling cycle.
14 so as to move by the pixel pitch pv.
To scan the object Q
You. One frame from sensor 53 for each sampling cycle
Is output. Focusing on one pixel g on the imaging surface S2,
5 samples out of N samples performed during scanning
In this case, effective light receiving data can be obtained. These 5 times
Of interest pixel g by interpolation operation
The optical axis of the slit light U passes through the object surface ag in the glare range
Timing (time center of gravity Npeak: reception of target pixel g)
The time at which the light amount becomes maximum) is obtained. In the example of FIG.
Is the time between the nth time and the previous (n-1) th time.
The received light amount is the largest in the ringing. At the timing I found
The irradiation direction of the slit light and the slit light for the pixel of interest
Position (coordinates) of the object Q based on the relationship with the incident direction of
Is calculated. As a result, the pixel pitch pv of the imaging surface is regulated.
Measurement with a higher resolution than the specified resolution becomes possible. The amount of light received by the target pixel g depends on the reflectance of the object Q.
Exist. However, the relative value of each received light amount of 5 samplings
The ratio is constant regardless of the absolute amount of received light. In other words, the object
Color shading does not affect measurement accuracy. In the measuring system 1 of the present embodiment, the three-dimensional
The received light data of 5 times for each pixel g of the sensor 53 by the camera 2
Is output to the host 3 as measurement data, and the host 3 measures
The coordinates of the object Q are calculated based on the data. 3D turtle
The output processing circuit 62 of FIG.
Responsible for generating measurement data in response. FIG. 15 is a block diagram of the output processing circuit 62.
FIG. 16 shows the reading range of the sensor 53. output
The processing circuit 62 controls the photoelectric conversion of each pixel g output from the sensor 53.
A / D converter that converts the conversion signal into 8-bit light receiving data
620, 4 frame delay memories connected in series
621 to 624, five effective received light data are stored
Memory banks 625A to 625A for receiving light reception data
Write the maximum frame number (sampling number) FN
Memory bank 625F for storing, comparator 62
6, a generator 627 indicating the frame number FN,
And address designation of the memory banks 625A to 625F.
And a memory control means (not shown). Each method
Molybank 625A-E is the number of sampling points for measurement
(That is, the same number of received light data as the number of effective pixels of the sensor 53)
Data storage capacity. Four frame delay memories 621 to 6
24, a data delay is applied to each pixel g.
At the same time, 5 frames of received light data are
25A to 25E. In addition,
The reading of one frame by the sensor 53 is performed by the imaging surface S2
As shown in Fig. 16 for speeding up,
Only a part of the effective light receiving area (band image) Ae of the image plane S2
Performed on the subject. The effective light receiving area Ae is the polarization of the slit light U.
The position is shifted by one pixel for each frame according to the direction. Real truth
In the embodiment, the number of pixels in the shift direction of the effective light receiving area Ae is
32. CCD image sensor
The method of reading out only a part is disclosed in Japanese Patent Laid-Open No. 7-174536.
It is disclosed in the gazette. The AD conversion unit 620 outputs 32
The received light data D620 for the line is serially arranged in the arrangement order of the pixels g.
Output to Al. Each frame delay memory 621-6
24 is an FI having a capacity of 31 (= 32-1) lines.
FO. The pixel of interest output from the AD converter 620
g of received light data D620 is delayed by two frames.
At that time, the comparator 626 activates the memory bank
625C stores the past received light data for the pixel of interest g.
Data D620. Delayed receiving data
Data D620 (output of frame delay memory 622)
If is larger than the past maximum value, the AD conversion at that time
Output of the section 620 and each frame delay memory 621 to 621
624 output to memory banks 625A-E respectively.
Stored, and the stored contents of the memory banks 625A to 625E are rewritten.
available. At the same time, the memory bank 625F contains
For the received light data D620 stored in the memory bank 625C,
The corresponding frame number FN is stored. That is, in the n-th (n <N) frame
When the amount of light received by the pixel of interest g is maximized, the memory
The data of the (n + 2) th frame is stored in the link 625A.
The (n + 1) th frame is stored in the memory bank 625B.
Is stored in the memory bank 625C.
The data of the eye frame is stored in the memory bank 625.
D stores the data of the (n-1) th frame.
The data of the (n-2) th frame is stored in the memory bank 625E.
And n is stored in the memory bank 625F.
You. Next, the operation of the three-dimensional camera 2 and the host 3
Will be described together with the measurement procedure. In the following,
The number of sampling points is 200 × 231. That is,
The number of pixels in the slit length direction on the image plane S2 is 231.
That is, the substantial number of frames N is also 200. The user (photographer) displays on the LCD 21.
Determine the camera position and orientation while viewing the color monitor image.
The angle of view. At that time, zooming if necessary
Perform the operation. In the three-dimensional camera 2, the color sensor 54
Aperture adjustment is not performed, and the electronic shutter function
The controlled color monitor image is displayed. This is the aperture
To make the incident light amount of the sensor 53
To do as much as possible. FIG. 17 shows the data of the three-dimensional camera 2.
FIG. 18 shows data and signals in the host 3.
FIG. 19 shows the relationship between each point of the optical system and the object Q.
Is shown. Angle-of-view selection operation by the user (zooming)
Of the variator unit 514 of the zoom unit 51 in accordance with
Movement is performed. Also, the movement of the focusing unit 512
Manual or automatic focusing is performed. Pho
During the focusing process, the approximate object distance d0 is measured.
It is. In response to such lens driving of the light receiving system,
The moving amount of the variator lens 422 on the light emitting side is shown in FIG.
Is calculated by an arithmetic circuit that does not
The movement control of the variator lens 422 is performed. The system controller 61 includes a lens controller.
Focusing via Troller 58 (FIGS. 3 and 5)
The output of the encoder 59A (the feeding amount Ed) and the zoom
The output (zoom step value fp) of the scaling encoder 60A
See in. In the system controller 61, distortion
Curvature aberration table T1, principal point position table T2, and image distance
With reference to the release table T3, the feed amount Ed and the zoom
Shooting condition data corresponding to the step value fp is output to the host 3
Is done. The shooting condition data here is based on the distortion parameters.
(Lens distortion correction coefficients d1, d2), front principal point position F
H and the image distance b. The front principal point position FH is zoomed
Expressed by the distance between the front end point F of the unit 51 and the front principal point H.
It is. Since the front end point F is fixed, the front principal point position FH
Thus, the front principal point H can be specified. The system controller 61 includes a semiconductor laser.
Output of the laser 41 (laser intensity) and deflection of the slit light U
(Scan start angle, scan end angle, deflection angular velocity)
You. This calculation method will be described in detail. First, an approximate
Assuming that a planar object exists at the object distance d0,
The projection angle is set so that the reflected light is received at the center of the
U. Pulse points for laser intensity calculation described below
Lighting is performed at the set projection angle. Next, the laser intensity is calculated. Laser intensity
Since the human body may be measured when calculating
Consideration for integrity is essential. First, the minimum strength LD
The pulse is turned on at min and the output of the sensor 53 is taken in.
Captured signal [Son (LDmin)] and appropriate level
Then, the ratio with the Typ is calculated, and the provisional laser intensity LD1 is calculated as follows: LD1 = LDmin × Styp / MAX [Son (LD
min)]. Subsequently, the pulse point is again set at the laser intensity LD1.
Lights, and captures the output of the sensor 53. Captured signal
[Son (LD1)] is at an appropriate level Type or
If the values are close, LD1 is determined as the laser intensity LDs.
In other cases, the laser intensity LD1 and MAX [Son (L
D1)] to set a temporary laser intensity LD1.
The output of the sensor 53 is compared with the appropriate level STyp. C
Until the output of the sensor 53 falls within the allowable range.
The provisional setting of the degree and the confirmation of suitability are repeated. The sensor 5
The capture of the output of No. 3 is performed for the entire surface of the imaging surface S2.
U. This is because the passive distance calculation by AF
It is difficult to estimate the light receiving position of the light U with high accuracy.
It is. The integration time of the CCD in the sensor 53 is 1
Field time (for example, 1/60 second) and the time of actual measurement
Longer than the integration time at. Therefore, pulse lighting is performed.
Thus, a sensor output equivalent to that at the time of measurement is obtained. Next, it is assumed that the projection angle and the laser intensity have been determined.
From the light receiving position of the slit light U by triangulation
The distance d is determined. Finally, the determined objective distance d
The deflection condition is calculated based on. When calculating deflection conditions
The rear side of the light receiving system, which is the reference point for distance measurement between the objectives d.
E in the Z direction (see FIG. 22) between the principal point H 'and the projection start point A
Consider the offset foff. Also, at the end in the scanning direction
The same measurable distance range d 'as in the central part.
Therefore, overscan of a predetermined amount (for example, for 8 pixels)
To do. Scan start angle th1, scan end angle th
2. The deflection angular velocity ω is expressed by the following equation. Th1 = tan ^-1 [(Β × pv × (np
/ 2 + 8) + L) / (d + doff)] × 180 / π th2 = tan ^-1 [(-Β × pv × (np / 2 + 8)
+ L) / (d + doff)] × 180 / π ω = (th1−th2) / np β: imaging magnification (= d / effective focal length freal) pv: pixel pitch np: effective number of pixels L in Y direction on imaging surface S2 : Baseline length Next under the condition calculated in this way
It shifts to light, and scanning (slit projection) of the object Q is performed.
5 per pixel obtained by the output processing circuit 52 described above.
Measurement data for frame (slit image data) D62
Is sent to the host 3. At the same time, the deflection conditions (deflection control data)
Device information D10 indicating the specifications of the
Is also sent to the host 3. Table 1 shows that the 3D camera 2
It summarizes the main data to be sent to the third party. [Table 1] As shown in FIG. 18, in the host 3,
Is the calculation of the center of gravity of the slit (# 31) and the correction of distortion
(# 32), calculation of camera line-of-sight equation (# 33), pick-up
Calculation of cut surface equation (# 34) and three-dimensional position calculation #
35 is executed, thereby obtaining 200 × 231 suns.
The three-dimensional position (coordinates X, Y, Z) of the pulling point is calculated
You. The sampling point is the camera line of sight (the sampling point and
Line connecting the side principal point H ') and the slit surface (sampling
This is the intersection with the slit light U irradiating the point (the optical axis plane). The time center of gravity Npeak of the slit light U (FIG. 1)
4) shows the received light data D (i) at each sampling.
And given by equation (3). Npeak = n + Δn (3) Δn = [− 2 × D (n−2) −D (n−1) + D (n +
1) + 2 × D (n + 2)] / ΣD (i) (i = n−2, n−1, n, n + 1, n + 2) or Δn = [− 2 × [D [n−2) −minD (i) )]-
[D (n-1) -minD (i)] + [D (n + 1)-
minD (i)] + 2 × [D (n + 2) -minD
(I)]] / ΣD (i) The minimum data minD (i) of the five received light data
By subtracting to obtain a weighted average, the effect of ambient light
Can be reduced. The camera line-of-sight equation is given by the following equations (4) and (5).
It is expressed by an equation. (U-u0) = (xp) = (b / pu) × [X / (Z-FH)] (4) (v−v0) = (yp) = (b / pv) × [Y / (Z -FH)] (5) b: Image distance FH: Front principal point position pu: Horizontal pixel pitch pv on the imaging surface: Vertical pixel pitch u on the imaging surface u: Horizontal pixel position u0 on the imaging surface: Center pixel position v in the horizontal direction on the imaging plane: vertical pixel position v0 on the imaging plane: center pixel position in the vertical direction on the imaging plane. The slit plane equation is expressed by equation (6).
You. (Equation 1) The geometric aberration depends on the angle of view. Distortion is almost center
Occurs in the object around the pixel. Therefore, the amount of distortion is
Expressed as a function of the distance from the center pixel. Here, the distance
Is approximated by a cubic function of The secondary correction coefficient is d1,
The correction coefficient is d2. The corrected pixel positions u ′ and v ′ are
It is given by equations (7) and (8). U ′ = u + d1 × t2 ² × (u−u0) / t2 + d2 × t2 ³ × (u−u0) / t2 (7) v ′ = v + d1 × t2 ² × (v−v0) / t2 + d2 × t2 ³ × (v−v0) / t2 (8) t2 = (t1) ^-2 t1 = (u-u0) ² + (V-v0) ² In the above equations (4) and (5), u
Substitute u 'for v and v' for v
Can determine the three-dimensional position taking into account the distortion
it can. For calibration, the electronic information
Proceedings of the IEICE Technical Meeting PRU91-113 [Camera positioning
Geometric correction of unnecessary images] Onodera / Kanaya, Electronic information
IEICE Transactions D-II vol. J74-D-II No.9 pp.1227-1
235, '91 / 9 [Range file based on 3D model of optical system
Indah's high-precision calibration method] Ueshiba, Yoshimi, Dai
There are detailed disclosures on islands, etc. The above embodiment is based on the measurement data D62.
That the host 3 is responsible for calculating the three-dimensional position
Is an operation for calculating the three-dimensional position of the three-dimensional camera 2
A function may be provided. Lookup table for 3D position
It is also possible to calculate by the rule method. Optical system 5 on the receiving side
0, replace the zoom unit 51 with an interchangeable lens
Therefore, the imaging magnification may be changed. Next, a method for obtaining three-dimensional data with no missing parts
The order will be described with reference to FIGS. S
Determine whether there is lack of continuity in lit image data
Automatically detects 3D data missing part automatically
Rotation of the object on the turntable and modification of the three-dimensional data
The return operation is performed automatically. The solid line in FIG.
The dashed line indicates the flow of data and the broken line indicates the flow of signal. 3D data
Preliminary scan to check slit image continuity
Followed by a main scan to collect 3D data
Done by At the time of preliminary scan, slit image data
Data is converted to a vector via the output terminal of the 3D camera
(# 36). Here, the vector transformation and the continuity determination will be described.
Will be described. In the vector conversion (# 36),
A method of thinning the set of pixel data
And plotting the resulting plotted data
The point and end point are defined, and the slit image data is
Vector conversion is performed. Then, determination of continuity (# 37)
In this, we use this vectorized data
The continuity is determined. The determination of continuity (# 37)
A diagram (vector) with all start and end points is the start point of each other
To see if it is continuously connected to the end point
If it is connected, the slit image data
It is determined that the line is continuous. In this case,
It is determined that there is no data loss. Also, continuity determination
In (# 37), the slit image is a broken line
If it is recognized that three-dimensional data is lost,
Refused. (All slit images in preliminary scan
(When images are continuous) In the flowchart of FIG.
The first scan of the preparatory scan is performed (S1). This
After that, the obtained slit image data
Vector conversion processing and continuous by vector conversion (# 36)
The continuity is determined by the sex determination (# 37). Communicating
In the determination of continuity (# 37), it is determined that the line is a continuous line.
When the result is obtained (YES in S2), the scan is completed.
It is determined whether or not it is the end angle, and if it is not the end angle (S
3; NO), and the next scan
The can angle is set (S4), and the second preliminary scan is performed
Is performed (S1). In this way, everything
Slit image data is continuous line
If determined (S2 and S3 for all scans)
YES in the projection side,
The scan start angle is set (S8), and the main scan is executed.
Then, three-dimensional data is obtained (S9). This place
In this case, there is no need to rotate
Three-dimensional data can be obtained. (Discontinuity of slit image during preliminary scan)
In the above case, the slip during the preliminary scan
When discontinuity of the image occurs (NO in S2),
A signal to stop the scanning operation is sent to the three-dimensional camera 2.
Then, the scanning operation is stopped. Next, turn the turntable 5
The rotating signal is sent to the turntable 5 so that the turntable 5 has a predetermined angle.
(S5). Then, at that scan angle
And a preliminary scan is performed (S6), and the slit image
Is determined, and if discontinuous (in S7,
NO), the rotation of the turntable 5 and the preliminary
Judgment of continuity is made. Angle setting of turntable 5
However, when the continuity of the slit image is obtained
(YES in S7), in the optical system 40 on the light emitting side
The scan start angle is set (S8), and the actual scan is performed.
A line is made. (Main Scan) In the main scan following the above,
May have discontinuities in the slit image. There
Then, the main scan is executed (S9), and the continuity is determined.
If it is not a continuous line (NO in S10),
The scan angle is stored in a predetermined memory (S12), and the next scan is performed.
The can angle is set (S13), and the continuity is determined (# 3).
If the result is a continuous line in (7), the process proceeds to (S10).
YES), set the next scan angle as it is
(S13), the main scan is executed at that angle (S9).
The main scan is executed up to the scan end angle. Next
Then, referring to the memory for storing the scan angle,
Whether slit image discontinuity occurs during main scan
And if there is no discontinuity, measure 3D data
Ends (NO in S14). (Generation of Discontinuity During Main Scan)
Refer to the memory that stores the scan angle
When there is a discontinuity of the slit image in the scan (S
14), all scans in which discontinuity has occurred
The rotation of the object on the turntable and the 3D data
Performs a data repair operation. First, the first stored in memory
The scan angle is set in the optical system 40 on the projection side.
(S15) Then, the turntable 5 is rotated by a predetermined angle (S16).
Scan is executed (S17), and continuity is determined (# 3).
At 7), the continuity of the slit image is determined (S1).
8). Discontinuous slit image at this scan angle
If (NO in S18), the rotation of the turntable is performed again.
And if scanning is continued (YE in S18)
S) Similarly, for the next scan angle,
A scan is performed to repair missing three-dimensional data.
If there is no longer a scan angle at which data should be restored (S1
(NO in 4) The measurement of the three-dimensional data ends. Up
In the procedure for obtaining three-dimensional data without missing parts described above,
Continuation of slit image in pre-scan and main scan
Discrimination of gender and rotation of turntable based on the discrimination result and 3
The operation of restoring dimensional data is all performed automatically. The present invention is not limited to the above configuration.
However, various modifications are possible. For example, three-dimensional
During data measurement, the operator of the measuring device
Even if it shifts to operation and returns to full automatic measurement again
Good. In addition, the preliminary scan and the main scan are set separately.
Without slitting the slit image
Judgment of continuity and rotation of turntable based on the judgment result
And a three-dimensional data restoration operation may be performed. As described above, according to the present invention, the slit
Vector conversion of the two-dimensional image of the reflected light
From the converted data, the continuity of the data,
3D data is automatically identified as missing, and there is no continuity
Place the subject so that continuity can be achieved
The rotating platform is rotated by the drive mechanism,
Receives light projection and the reflected light of the projection light, thereby causing loss
3D data without parts can be collected automatically
You. Thus, the shape of the object may be complicated.
From the position where the detection light is illuminated.
When the detection light cannot be projected or when viewed from the imaging position
The reflected light from the shaded area cannot be measured.
However, it is possible to deal with it automatically.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a configuration diagram of a measurement system according to the present invention. FIG. 2 is a diagram illustrating an appearance of a three-dimensional camera. FIG. 3 is a block diagram illustrating a functional configuration of a three-dimensional camera. FIG. 4 is a schematic diagram showing a configuration of a light projecting lens system. FIG. 5 is a schematic diagram of a zoom unit for receiving light. FIG. 6 is a schematic diagram of a beam splitter. FIG. 7 is a diagram showing a light receiving wavelength of a sensor for measurement. FIG. 8 is a diagram illustrating a light receiving wavelength of a color sensor for monitoring. FIG. 9 is a perspective view schematically illustrating a two-axis adjustment mechanism of the optical unit. FIG. 10 is a front view of the upper part of the optical unit as viewed from the direction of arrow KA. FIG. 11 is a right side view of the upper part of the optical unit as viewed from the direction of arrow KB. FIG. 12 is a bottom view of the same optical unit as viewed from the direction of arrow KC. FIG. 13 is a diagram for explaining a method of adjusting a two-axis adjustment mechanism of the optical unit. FIG. 14 is a principle diagram of calculation of a three-dimensional position in the measurement system. FIG. 15 is a block diagram of an output processing circuit. FIG. 16 is a diagram showing a reading range of a sensor. FIG. 17 is a diagram showing a data flow in the three-dimensional camera. FIG. 18 is a diagram showing a data flow in the host. FIG. 19 is a diagram showing a relationship between each point of the optical system and an object. FIG. 20 is a flowchart illustrating a preliminary scan, a main scan, and a rotation operation of the turntable. FIG. 21 is a diagram showing an outline of a conventional slit light projection method. FIG. 22 is a diagram for explaining the principle of measurement by a conventional slit light projection method. [Description of Signs] 1 Measurement system 2 3D camera (3D input camera) 3 Host 3a CPU (vector conversion means, determination means for determining continuity) 5 Turntable 40 Optical system (projection means) 50 Optical system (light reception) Means) AX1 First rotation axis AX2 Second rotation axis U Slit light (detection light) Q Object (subject) # 36 Vector conversion # 37 Determination of continuity

   ────────────────────────────────────────────────── ─── Continuation of front page    F term (reference) 2F065 AA04 DD00 DD06 DD12 EE00                       EE08 FF01 FF02 FF09 GG06                       GG08 GG22 HH05 JJ03 JJ26                       LL04 LL06 LL09 LL10 LL13                       LL20 LL21 LL26 LL30 LL46                       MM16 NN02 NN17 PP05 PP13                       QQ00 QQ03 QQ17 QQ24 QQ29                       QQ32 SS02 SS13                 5B057 AA20 BA02 DA07 DB03 DB06                       DB09 DC07 DC09

Claims (1)

1. A projecting means for projecting slit light to optically scan a subject, and receiving reflected light of the slit light projected on the subject by the projecting means from the subject. Light receiving means, a turntable for mounting the subject,
A three-dimensional measuring device comprising a driving mechanism for rotating the turntable and measuring the shape of the subject in a non-contact manner; a means for performing vector conversion on two-dimensional image data consisting of slit-like reflected light received by the light receiving means And determining means for determining the continuity of the vector-converted image data. Each time the slit light scans, the obtained image data is vector-converted by the vector-converting means. The continuity of the data is determined by the determination means, and when it is determined that the continuity is lacking, the subject is rotated by rotating the turntable by the driving mechanism at the scanning position of the slit light, and again. A three-dimensional measuring apparatus for projecting the slit light and receiving reflected light of the projected light.