JP2007315777A - Three-dimensional shape measurement system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a three-dimensional shape measurement system which can measure the three-dimensional shape of an object together with its density distribution at a high speed. <P>SOLUTION: All the brightness of the pixels of an image frame b2 photographed by a camera b1 and an image frame c2 photographed by a camera c1 in which (a) coordinates 411 within an area 401 having density within a density level 2 of a fog-like body in the real world RW is imaged is the brightness corresponding to the density level 2. A (d) voxel 421 on virtual space VW corresponding to the coordinates 411 is therefore set as a voxel having the density level 2 as an attribute within the fog-like body. The brightness of the pixels of the image frame b2 and the image frame c2 in which (a) coordinates 412 within an area 402 having density within a density level 1 of the fog-like body in the RW is the brightness corresponding to the density level 1. A (d) voxel 422 on the VW corresponding to the coordinates 412 is therefore set as a voxel having the density level 1 as an attribute within the fog-like body. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a technique for measuring a three-dimensional shape of an object.

As a technique for measuring the three-dimensional shape, the surface of the object is scanned with slit light, and the optical cutting method is used to measure the three-dimensional shape of the object from the shape of the slit light that appears in each image obtained by photographing the object. Is known (for example, Patent Document 1).
Japanese Unexamined Patent Publication No. 09-210646

  Now, according to the light cutting method, since it is necessary to scan the entire surface of the object with slit light every time a three-dimensional shape is measured, high-speed three-dimensional shape measurement cannot be performed. For this reason, it is difficult to appropriately measure the three-dimensional shape of an object that deforms at a relatively high speed and the transition of the change in the three-dimensional shape with the measurement technique based on the light cutting method. In addition, when the object to be measured is an object whose mist or smoke density distribution changes, the density distribution or change thereof cannot be measured.

  Therefore, an object of the present invention is to measure the three-dimensional shape of an object together with its density distribution at a higher speed.

  In order to achieve the above object, the present invention provides a three-dimensional shape measurement system for measuring a three-dimensional shape of an object to be measured, a plurality of cameras for photographing the object to be measured, and a three-dimensional image photographed by the plurality of cameras. For each of the coordinates in the space, the pixel of the pixel in the image captured by the plurality of cameras on which the coordinates are projected to any of the pixel range for each density level, which is a pixel value range determined for each density level. By calculating whether all of the values are included and setting the density level corresponding to the calculated pixel range for each density level as the density level of the coordinates, the three-dimensional shape for each density level of the measurement object is obtained. It includes a three-dimensional shape measuring means for measuring.

  In order to achieve the above object, the present invention provides a three-dimensional shape measurement system for measuring a three-dimensional shape of a measurement object, which is photographed by a plurality of cameras that photograph the measurement object and the plurality of cameras. Of the coordinates in the three-dimensional space, the pixel value of the pixel on which the coordinates in the image photographed by the camera are projected is set in all the images photographed by the plurality of cameras. A three-dimensional shape measuring means for measuring the three-dimensional shape of the measured object by a visual volume intersection method in which the coordinates of the pixel values within the range of values are the coordinates inside the measured object, and a plurality of different said For the extraction target pixel value range, by causing the 3D shape measurement means to measure the 3D shape of the measurement object, the density level of the measurement object is measured for each of a plurality of different density levels. Is constructed by including a three-dimensional density distribution measuring means for measuring the three-dimensional shape of a region having a density corresponding to.

  According to the three-dimensional shape measurement systems such as these, for each density level of the object to be measured from images obtained by simultaneously photographing the object to be measured with a plurality of cameras using a visual volume intersection method that does not require scanning of slit light or the like. The three-dimensional shape of the area can be measured. Therefore, according to these three-dimensional shape measurement systems, the three-dimensional shape and three-dimensional density distribution of the measurement object can be measured at high speed, and even for objects whose shape and density distribution change relatively quickly, It becomes possible to appropriately measure the transition of the original shape and the three-dimensional density distribution and the change of the three-dimensional shape and the three-dimensional density distribution.

  In each of the above three-dimensional shape measurement systems, the object to be measured may be a sprayed object.

  As described above, according to the present invention, the three-dimensional shape of an object can be measured together with its density distribution at a higher speed.

Hereinafter, embodiments of the present invention will be described.
FIG. 1 shows a configuration of a three-dimensional shape measurement system according to the present embodiment.
As shown in the figure, the three-dimensional shape measurement system includes a plurality of sets of a camera 1 and a video input interface 2 that captures an image from the camera 1. The three-dimensional shape measurement system includes a three-dimensional shape modeling unit 3, a three-dimensional shape memory 4, a rendering unit 5, a display device 6, and an input device 7.

  Here, the cameras 1 shoot the measurement object in synchronism with each other at the same frame period, and output the image frames captured for each frame period to the video input interface 2 that forms a set. The video input interface 2 stores image frames sequentially input from the paired cameras 1 in a frame memory provided therein.

  Next, the three-dimensional shape modeling unit 3 reads out each image frame photographed by the plurality of cameras 1 at the same time stored in the frame memory of each video input interface 2 to obtain the three-dimensional shape of the object to be measured. Then, measurement is performed by the visual volume intersection method, and the data of the three-dimensional model having the measured three-dimensional shape is stored in the three-dimensional shape memory 4 as three-dimensional model data.

The rendering unit 5 displays the three-dimensional shape of the object to be measured on the display device 6 in a form designated by the operator via the input device 7 based on the three-dimensional shape model data stored in the three-dimensional shape memory 4. To display.
Note that such a three-dimensional measurement system may be configured using an electronic computer having a general-purpose configuration. In this case, a plurality of video input interfaces 2, a display device 6 and an input device 7 each connected to a camera are connected to an electronic computer, and the aforementioned three-dimensional shape modeling unit 3, three-dimensional shape memory 4 and rendering unit 5 are connected. Is implemented in an electronic computer as a process embodied as a storage resource on the electronic computer by executing a predetermined program prepared in advance.

Hereinafter, the details of the three-dimensional shape measurement operation of such a three-dimensional shape measurement system will be described by taking the application to the measurement of the three-dimensional shape of the sprayed mist as an example.
FIG. 2 shows an arrangement example of the camera 1 of the three-dimensional shape measurement system when such measurement is performed.
In the illustrated example, the mist 221 sprayed downward from the nozzle 220 is measured. In this example, nine units arranged on the circumference of a circle centering on the injection axis of the nozzle below the nozzle 220. The direction of the ejection port of the nozzle 220 is taken from below with the camera 1. In addition, an illumination device 201 that emits illumination light 202 that illuminates the direction of the nozzle 220 from below, and a transparent plate 210 for protecting the camera disposed between the nozzle 220 and the camera 1 are provided.

  Next, the 3D shape modeling process performed by the 3D shape modeling unit 3 will be described. The three-dimensional shape modeling process realizes the above-described operation of measuring the three-dimensional shape of the object to be measured by the three-dimensional shape modeling unit 3 and storing the three-dimensional model data in the three-dimensional shape memory 4. .

FIG. 3 shows the procedure of this three-dimensional modeling process.
As shown in the figure, in this process, when measurement is started (step 302), the following process is repeated until the measurement is completed (step 340).
That is, first, the current time is acquired as t (step 304). Also, the latest image frame stored in the frame memory of each video input interface 2 is acquired as the image frame at the time t (step 306). Then, the coordinates (x, y, z) of the measurement target real space RW are sequentially set as processing target coordinates (steps 308, 310, 312, 326, 332, 328, 334, 330, 336), and the processing target coordinates are objects. A visual volume intersection region calculation process (step 314 to step 324) is performed for discriminating between the internal coordinates and the external coordinates of the object by the visual volume intersection method. In this visual volume intersection region calculation process, when the processing target coordinates are coordinates inside the object, a process of measuring the density level of the processing target coordinates of the object is also performed.

That is, in this visual volume intersection region calculation process (steps 314 to 324), each camera 1 acquired from each video input interface 2 in step 306 is set with n as the number of density decompositions and s as the luminance step width for decomposition. In each captured image frame, the maximum i for which all the luminances D (x, y, z) of the pixels onto which the processing target coordinates (x, y, z) are projected is equal to or higher than Thi, i is changed from n to n. While decreasing to 1 (steps 314, 320, 324), calculation is made (step 316). However,
Th i = reference threshold + {(i−1) × s}
And
Here, the pixel on which the processing target coordinates (x, y, z) are projected in the image frame refers to an obstacle in capturing the processing target coordinates (x, y, z) of the camera 1 that captures the image frame. In the case where there is no shielding object, the pixel in the image frame in which an object existing at the processing target coordinates (x, y, z) is reflected. The density decomposition number is set in advance when the density of an object is measured in n stages from density level 1 to density level n. In addition, the resolution luminance step width s is a value obtained by dividing the difference between the luminance obtained when the portion having the maximum density of the object is photographed and the luminance obtained when the portion having the minimum density of the object is photographed by n. Set. In addition, as the reference threshold value, a luminance value of a pixel that has captured an object having the minimum density within the density level 1 is set in advance. Note that, when shooting is performed as shown in FIG. 2, as the density inside the mist 221 increases, the luminance of the pixel that has shot the position also increases.

  If the maximum i can be obtained, the voxel in the coordinates (x, y, z) of the virtual space VW (t) at time t is set as the voxel inside the object, and the obtained maximum i is set to the voxel of the voxel. The density level attribute is set (step 318). Further, when the maximum i cannot be obtained, that is, any of the luminances D (x, y, z) of the pixels onto which the processing target coordinates (x, y, z) are projected is less than the reference threshold value. If so, the voxel of the coordinates (x, y, z) in the virtual space VW (t) at time t is set as a voxel outside the object (step 322).

Here, FIG. 4 shows a processing example by the above-described visual volume intersection region calculation processing (steps 314 to 324).
In the illustrated example, the number of density decompositions is 2, and the density in the region 402 having the density in the density level 1 existing in the real world RW and the density in the density level 2 (density level 2> density level 1) as shown in a. The mist having the region 401 is photographed from different directions by the camera 1 shown in b1 and the camera 1 shown in c1. The image frame shown in b2 is shot by the camera 1 shown in b1, and the image frame shown in c2 is shot by the camera 1 shown in c1.

  In this case, the luminance of the pixel in the image frame b2 on which the coordinates 411 in the region 401 having a density within the density level 2 of the mist in the real world RW shown in a is projected, and c2 on which the coordinates 411 are projected. Since the luminance of all the pixels of the image frame is equal to or higher than the luminance (reference threshold + s) corresponding to the minimum density within the density level 2, it corresponds to the coordinate 411 in the real world RW shown in a. The voxel 421 on the virtual space VW shown is set as a voxel having density level 2 as an attribute inside the object.

  Next, the luminance of the pixel in the image frame b2 on which the coordinates 412 in the region 402 having the density within the density level 1 of the mist in the real world RW shown in a is projected, and the coordinates 212 on which the coordinates 412 are projected. Since the luminances of the pixels of the image frame are not all equal to or higher than the luminance corresponding to the minimum density within the density level 2 (reference threshold value + s), d corresponds to the coordinate 412 in the real world RW shown in a. The voxel 422 on the virtual space VW shown is not set as a voxel having density level 2 as an attribute inside the object. However, since the luminances of these two pixels are all equal to or higher than the luminance (reference threshold value) corresponding to the minimum density within the density level 1, the luminance is indicated by d corresponding to the coordinates 412 in the real world RW indicated by a. The voxel 422 on the virtual space VW is set as a voxel having density level 1 as an attribute inside the object.

  Further, the luminance of the pixel in the image frame b2 on which the coordinates 413 in the region 402 having the density within the density level 1 of the mist in the real world RW shown in a is projected, and the c2 on which the coordinates 413 are projected. The luminance of the pixel in the image frame of c2 is equal to or higher than the luminance corresponding to the minimum density within the density level 2 (reference threshold + s), but the luminance of the pixel in the image frame of b2 is There is a relationship that does not exceed the luminance (reference threshold + s) corresponding to the minimum density in density level 2. Therefore, the voxel 423 on the virtual space VW indicated by d corresponding to the coordinate 413 in the real world RW indicated by a is not set as a voxel having density level 2 as an attribute inside the object. However, since the luminances of these two pixels are all equal to or higher than the luminance (reference threshold value) corresponding to the minimum density within the density level 1, the luminance is indicated by d corresponding to the coordinates 413 in the real world RW indicated by a. A voxel 423 on the virtual space VW is set as a voxel having density level 1 as an attribute inside the object.

  Finally, the brightness of the pixel of the image frame b2 where the coordinates 414 outside the mist in the real world RW shown in a are reflected, and the brightness of the pixel of the image frame c2 where the coordinates 414 are reflected are all Since the luminance corresponding to the minimum density in the density level 2 (reference threshold value + s) does not exceed the luminance (reference threshold value) corresponding to the minimum density in the density level 1, the luminance does not exceed a. The voxel 424 in the virtual space VW shown in d corresponding to the coordinates 414 in the real world RW shown in FIG.

  Now, returning to FIG. 3, when the above processing is completed for each coordinate (x, y, z) of the real space RW to be measured, each of the objects inside the object in the virtual space VW (t) at time t The voxel data is stored in the three-dimensional shape memory 4 as the three-dimensional model data measured at time t (step 338).

The three-dimensional shape modeling process has been described above.
Returning to FIG. 1, the rendering unit 5 converts the three-dimensional shape of the measurement object via the input device 7 based on the three-dimensional model data stored in the three-dimensional shape memory 4 as described above. The information is displayed on the display device 6 in a form designated by the operator.
FIG. 5 shows a display example on the display device 6 by the rendering unit 5 as described above, and a1 is data of each voxel inside the object represented by the three-dimensional model data measured at time t = t1. Based on the above, a two-dimensional image in which a three-dimensional shape of an object that is a mist-like object is projected is generated, and the three-dimensional shape of the mist-like object is pseudo-three-dimensionally displayed by the generated two-dimensional image. Similarly, b1 is a pseudo three-dimensional display of the three-dimensional shape of the mist based on the data of each voxel inside each object represented by the three-dimensional model data measured at time t = t2. .

  Here, the pseudo three-dimensional display of the three-dimensional shape of the mist as shown in a1 or b1 corresponds to the density level set as an attribute for each voxel inside the object represented by the three-dimensional model data. In addition, the two-dimensional image generated by projecting each voxel onto a projection plane determined according to the viewpoint set by the operator is displayed after giving transparency and color. The transparency given to each voxel increases as the density level decreases.

By doing in this way, the shape of a mist-like body can be displayed in the form which can grasp | ascertain three-dimensional density distribution directly.
Next, a2 indicates the density distribution in the cross section 501 of a1 based on the density level set in each voxel, and b2 sets the density distribution in the cross section 502 of b1 in each voxel. It is displayed based on the density level.
Note that the rendering unit 5 automatically advances the time t, and the pseudo tertiary of the three-dimensional shape of the mist based on the data of each voxel inside the object represented by the three-dimensional model data measured at the time t. By performing the original display as shown by a1 and b1, the change in the three-dimensional shape of the shape may be displayed as an animation.

The embodiment of the present invention has been described above.
Incidentally, the three-dimensional modeling process performed by the three-dimensional modeling unit described above may be performed as shown in FIG.
That is, in this three-dimensional modeling process, when measurement is started (step 602), the following process is repeated until the measurement is completed (step 616).
That is, first, the current time is acquired as t (step 604). Also, the latest image frame stored in the frame memory of each video input interface 2 is acquired as the image frame at the time t (step 606). And for each i between 1 and n (steps 608, 614, 616),
Th i = reference threshold + {(i−1) × s}
As above, Thi or more is set as the object internal pixel luminance range (step 610), i-th density level three-dimensional model data is created by the view volume intersection method, and stored in the three-dimensional shape memory 4 (step 612). More specifically, in step 612, among the coordinates of the physical space RW, the coordinates of the pixels onto which the coordinates are projected in the image frames captured by the cameras 1 are all within the object internal pixel luminance range. The corresponding voxel on the virtual space VW is a voxel inside the object of the i-th density level. Here, the object of the i-th density level is an object whose inside has a density equal to or higher than Thi.

  By performing the three-dimensional modeling process in this way, three-dimensional model data of an object having a density equal to or higher than the corresponding density level corresponding to each density level is created at each time point. Therefore, the 3D shape represented by each 3D model data can be displayed individually or in combination, or the 3D shape generated by the operation between the 3D shapes represented by each 3D model data can be displayed intuitively. Thus, the shape of the mist can be displayed in a form in which the density distribution can be grasped.

Further, in the above embodiment, the application in the case where the sprayed mist is the measurement target has been described as an example. However, the three-dimensional measurement system according to the present embodiment is not limited to the mist, but smoke and milk. The same applies to turbid and suspended objects.
In addition, in the above, the application to the mist-like body in which the luminance that is photographed as the density is high is taken as an example, but in the case of applying to the object that the luminance that is photographed as the density is high, What is necessary is just to invert and apply the magnitude level of the density level in the above embodiment.

It is a block diagram which shows the structure of the three-dimensional shape measurement system which concerns on embodiment of this invention. It is a figure which shows the example of camera arrangement | positioning at the time of the measurement by the three-dimensional shape measurement system which concerns on embodiment of this invention. It is a flowchart which shows the three-dimensional modeling process which concerns on embodiment of this invention. It is a figure which shows the process example of the three-dimensional modeling process which concerns on embodiment of this invention. It is a figure which shows the example of a display of the three-dimensional shape measurement system which concerns on embodiment of this invention. It is a flowchart which shows the other example of the three-dimensional modeling process which concerns on embodiment of this invention.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 ... Camera, 2 ... Video input interface, 3 ... Three-dimensional shape modeling part, 4 ... Three-dimensional shape memory, 5 ... Rendering part, 6 ... Display apparatus, 7 ... Input device.

Claims (5)

A three-dimensional shape measurement system for measuring a three-dimensional shape of a measurement object,
A plurality of cameras for photographing the object to be measured;
For each of the coordinates in the three-dimensional space photographed by the plurality of cameras, the coordinates are projected onto any of the pixel ranges for each density level that are pixel value ranges determined for each density level. By calculating whether all the pixel values of the pixels in the image photographed by the camera are included and setting the density level corresponding to the calculated pixel range for each density level as the density level of the coordinates, the measurement target A three-dimensional shape measuring system comprising: a three-dimensional shape measuring means for measuring a three-dimensional shape for each density level of the body.
A three-dimensional shape measurement system for measuring a three-dimensional shape of a measurement object,
A plurality of cameras for photographing the object to be measured;
Of the coordinates in the three-dimensional space photographed by the plurality of cameras, the pixel values of the pixels onto which the coordinates in the image photographed by the camera are projected are all of the images photographed by the plurality of cameras. A three-dimensional shape for measuring the three-dimensional shape of the measured object by a visual volume intersection method using the coordinates within the set extraction target pixel value range as coordinates inside the measured object Measuring means;
For each of a plurality of different extraction target pixel value ranges, the three-dimensional shape measurement unit measures the three-dimensional shape of the measurement object, whereby the measurement object is measured for each of a plurality of different density levels. A three-dimensional shape measurement system comprising: a three-dimensional density distribution measuring unit that measures a three-dimensional shape of a region having a density corresponding to the density level.
The three-dimensional shape measurement system according to claim 1 or 2,
The three-dimensional shape measurement system, wherein the object to be measured is a sprayed object.
A computer program that is read and executed by an electronic computer,
The electronic computer,
For each of the coordinates in the three-dimensional space photographed by a plurality of cameras, the plurality of cameras on which the coordinates are projected to any one of the pixel range for each density level, which is a pixel value range determined for each density level A calculation means for calculating whether all of the pixel values of the pixels in the image taken with
The density level corresponding to the pixel range for each density level calculated by the calculation means is set to the density level of the coordinates, thereby measuring the three-dimensional shape for each density level of the measurement object. A computer program which functions as a shape measuring means.
A computer program that is read and executed by an electronic computer,
The electronic computer,
Of the coordinates in the three-dimensional space photographed by a plurality of cameras, the pixel values of the pixels onto which the coordinates in the image photographed by the camera are projected are all in the images photographed by the plurality of cameras. 3D shape measurement for measuring the 3D shape of the measured object by the visual volume intersection method using the coordinates within the set extraction target pixel value range as the coordinates inside the measured object Means,
For the plurality of different extraction target pixel value ranges, the density of the measurement object is measured for each of a plurality of different density levels by causing the three-dimensional shape measurement unit to measure the three-dimensional shape of the measurement object. A computer program that functions as a three-dimensional density distribution measuring unit that measures a three-dimensional shape of an area having a density corresponding to a level.