JP5079402B2 - Method and apparatus for measuring a three-dimensional profile of a light emitting object - Google Patents

Description

  The present invention relates to a method for measuring the shape of an object that self-lumines when heated and a device for measuring the shape. Specifically, for example, the present invention relates to a method and apparatus for non-contact measurement of surface three-dimensional profile data of a refractory in a high-temperature heat-resistant container storing molten metal or a refractory in a heating furnace.

  In order to repair the refractory that is the inner lining of the container that holds the molten metal or the inner surface of the heating furnace, it is necessary to measure the three-dimensional profile of the surface of the refractory and to determine its degree of wear. is there. In particular, grasping the remaining thickness of the refractory is strongly desired for improving productivity because repair work can be performed without missing the machine while maintaining safety in production. Therefore, it is generally performed to estimate the remaining thickness of the surface of the refractory having such a high temperature by a non-contact optical measurement method. However, since the object, that is, the refractory, has a high temperature, the surface itself emits light. For this reason, when the remaining thickness of the refractory is measured by a non-contact optical measurement method, there is a problem that a measurement error is likely to occur or measurement is partially impossible.

  Therefore, as an optical measurement method using laser light or the like, there is a measurement method using the fact that the energy density of a unit area is increased by spot irradiation of laser light or the like. Many systems that can measure the shape of an object with high accuracy in a non-contact manner from a relatively long distance by using this optical measurement method using laser light have been proposed and put into practical use. Many of such optical measurement methods using laser light or the like use a measurement method using the time of flight of light or a triangulation method using a spot irradiation bright spot. However, among these methods, when measuring the entire surface of an object such as a refractory by the measurement method using the time of flight of light, the measurement point by the laser beam is set to the entire surface of the object (refractory). Must be scanned against. Therefore, a highly accurate scanning mechanism is required, which is inefficient and tends to be expensive.

  On the other hand, the stereo image method, which is a kind of triangulation method, has an advantage that it can quickly measure an object with a two-dimensional image sensor. However, due to the uncertainties of matching that extracts matching points in two images obtained by the left and right imaging devices, it has been made unsuitable for high-precision measurement of the surface of a refractory. That is, in the stereo image method, feature points and patterns of shadow images by ambient light are used as matching points. However, there are many similar feature points and patterns on the surface of the refractory. For this reason, in the stereo image system, the above-described uncertainties of matching occur.

  On the other hand, Patent Document 1 discloses a system that uses a rotating slit light in a stereo image system. In this system, feature points are generated in an active and limited manner, so that the uncertainty of feature point matching in the stereo image method is improved. In the light cutting method, the accuracy of the turning angle of the slit light has a great influence on the measurement accuracy in the system, but this effect can be reduced in this system.

  Patent Document 2 discloses a technique for correcting a captured image obtained with laser slit light using a captured image obtained with ambient light. In this technique, relative unevenness of an object is obtained from a displacement curve of laser slit light applied obliquely. Then, three-dimensional quantitative shape information on the object is obtained from the information on the relative irregularities of the obtained object and information on two-dimensional known dimensions such as joints obtained by ambient light.

  By the way, if the output of laser light is high, there is a danger to the human body, especially the visual system, so when using high-power laser light, cover the target area so that the laser light and reflected light do not leak. It is obliged to cover and ensure safety. The work covered with a cover is a heavy burden as a work between operations, and the loss of the operation stop is large. For this reason, it is desired to perform measurement with a low-power laser beam that can omit the work of covering the target area with a cover. However, when a low-power laser beam is used when measuring the shape of a light-emitting object using an optical measurement method, the measurement accuracy and efficiency of the conventional technology described above are reduced. Therefore, it has been a big problem to eliminate the work of covering the target area with the cover and to suppress the decrease in measurement accuracy and efficiency.

JP 61-31909 A JP 2002-90124 A

The system disclosed in Patent Document 1 is assumed to be used on the surface of an object that does not emit light. Therefore, if this system is used as it is for the surface of an object to emit light, the following problems arise.
First, when slit light of non-laser light (light other than laser light) is used as the light source, the S / N ratio (the strength of the signal component with respect to noise) decreases, making measurement impossible, The efficiency is significantly reduced. On the other hand, when laser light is used as a light source, a large output light source is required to obtain sufficient luminance in a line-shaped projection area that is larger than the spot light projection area. Therefore, there was a limit to measurement within safety standards.

Further, since the shape is perpendicular to the shape perpendicular to the slit light, the shape recognition resolution is low. In particular, it is difficult to measure if cracks and groove-like shapes that are important as defects are arranged parallel to the slit light.
Therefore, the system disclosed in Patent Document 1 is not suitable for practical use of high temperature refractory profile measurement.

  In the technique disclosed in Patent Document 2, measurement of a reference shape such as a joint is performed by using ambient light. However, since the measurement of the three-dimensional profile by the ambient light is not performed, the quantitativeness of the measurement is lost when the features such as the joints cannot be found due to the attached matter. Of course, the technique described in Patent Document 2 is applied to an object that does not have a reference shape (such as a joint) whose dimensions can be found by two-dimensional image processing, such as the surface of a refractory constructed with a sprayed body. It is impossible.

  In view of the above circumstances, an object of the present invention is to provide a highly accurate three-dimensional profile measuring method and measuring apparatus capable of estimating the remaining thickness of a high-temperature object that emits light (for example, a refractory). To do. This is also a measurement method and a measurement device that can measure quickly and ensure safety to surrounding workers so that the interval between operations can be used regardless of the surface properties of the refractory.

The present invention has been made to solve the above-mentioned problems, and the gist thereof is as follows.
(1) Stereo measurement using a pair of stereo cameras installed at a predetermined interval to image a light emitting object and obtain a three-dimensional profile. Laser slit light having a constant wavelength between 400 and 680 nm is used as the object. In the measurement method of a three-dimensional profile of a light emitting object, which is measurement performed by irradiation,
The imaging mode is switched to an imaging mode (hereinafter referred to as laser beam imaging mode) in which the laser slit light is irradiated by mode switching means,
After the imaging mode is switched to the laser light imaging mode, irradiate the laser slit light on the surface of the object,
Taking an image of the surface of the object irradiated with the laser slit light by the stereo camera pair,
By performing stereo image processing on the captured image, first three-dimensional profile data based on the laser slit light is obtained,
The imaging mode is switched to an imaging mode (hereinafter referred to as ambient light imaging mode) in which the laser slit light is not irradiated by the mode switching means,
After the imaging mode is switched to the ambient light imaging mode, a shadow generated on the surface of the object is generated by the stereo camera pair based on the ambient light and the self-light emitted from the surface of the object. Take an image containing
Stereo image processing is performed on the captured image to obtain second three-dimensional profile data based on the ambient light and the self-light emission,
The position of the second three-dimensional profile data is corrected with reference to the first three-dimensional profile data, and then the second three-dimensional profile data is merged with the first three-dimensional profile data. The three-dimensional profile data of the whole object is obtained by the above method, and the method for measuring the three-dimensional profile of the object to emit light is provided.

  (2) When obtaining the first three-dimensional profile data by the stereo image processing, the laser slit light irradiated through the light receiving wavelength selection means that preferentially passes the reflected light of the laser slit light The method for measuring a three-dimensional profile of a light-emitting object according to (1), wherein an image of the surface of the object is captured.

( 3 ) In the first three-dimensional profile data and the second three-dimensional profile data, a deviation amount between the two data is evaluated by a predetermined method,
If pre said deviation amount from the limits set in large, the entire second three-dimensional profile data after moving parallel to the optical axis of the stereo camera pair, perpendicular to the optical axis of the stereo camera pair and performing that you rotate moving the entire second three-dimensional profile data as the rotation axis of two mutually orthogonal axes, modifying said second three-dimensional profile data so that the deviation amount becomes minimum As a modified second 3D profile data,
The modified second three-dimensional profile data is used in place of the original second three-dimensional profile data and fused with the first three-dimensional profile data to obtain three-dimensional profile data of the entire object. The method for measuring a three-dimensional profile of a light-emitting object according to the above (1) or (2) .

( 4 ) The position of each point of the first three-dimensional profile data and the second three-dimensional profile data is compared in the optical axis direction view of the stereo camera pair, and the overlapping point or a predetermined distance A closer point pair is a neighbor point pair,
The distance in the optical axis direction of the neighboring point pair as a deviation distance,
Obtaining the deviation distance for at least four or more neighboring point pairs arbitrarily selected from among neighboring point groups that are a set of neighboring point pairs in the entire first and second three-dimensional profile data;
The method for measuring a three-dimensional profile of a light-emitting object according to ( 3 ), wherein a square sum of the obtained deviation distances is obtained as the deviation amount.
(5) the so deviation amount becomes minimum, after the second three-dimensional profile data across moved parallel to the optical axis of the stereo camera pair, which translation modifications three-dimensional profile data,
It passes through the center of gravity of the translation correction three-dimensional profile data or the point having the arithmetic mean of the coordinate values of each point of the translation correction three-dimensional profile data as the coordinate value, and on the optical axis of the stereo camera pair Rotating and moving the translation correction three-dimensional profile data with an arbitrary one axis orthogonal to the rotation axis, this as one axis rotation movement correction three-dimensional profile data,
The one-axis rotational movement correction three-dimensional profile data with the axis passing through the center of gravity of the one-axis rotational movement correction three-dimensional profile data and orthogonal to the optical axis of the stereo camera pair and the arbitrary one axis as the rotation axis The object to emit light according to ( 3 ) or ( 4 ), wherein the object is rotated again so that the amount of deviation is minimized, and this is used as the modified second three-dimensional profile data. 3D profile measurement method.
( 6 ) The stereo camera pair includes a first stereo camera pair that captures an image in the laser light imaging mode and a second stereo camera pair that captures an image in the ambient light imaging mode. The method for measuring a three-dimensional profile of a light-emitting object according to any one of (1) to ( 5 ).

( 7 ) In a three-dimensional profile measuring apparatus that irradiates an object to emit light with laser slit light having a constant wavelength of 400 to 680 nm and performs stereo measurement using a stereo camera pair installed at a predetermined interval.
Mode switching means for switching an imaging mode between an imaging mode for irradiating the laser slit light (hereinafter referred to as a laser light imaging mode) and an imaging mode for not irradiating the laser slit light (hereinafter referred to as an ambient light imaging mode);
Laser slit light irradiation means for irradiating the surface of the object with the laser slit light;
The image of the surface of the object irradiated with the laser slit light, and the image including the shadow generated on the surface of the object based on the ambient light and the self-light emitted from the surface of the object. A pair of stereo cameras for imaging
The image of the surface of the object irradiated with the laser slit light is image-processed to obtain first three-dimensional profile data, and a shadow generated on the surface of the object based on the ambient light and the self-light emission. Stereo image processing means for obtaining second three-dimensional profile data by image processing an image including:
The position of the second three-dimensional profile data is corrected with reference to the first three-dimensional profile data, and then the second three-dimensional profile data is merged with the first three-dimensional profile data. By means of data fusion means for obtaining three-dimensional profile data of the entire object,
A device for measuring a three-dimensional profile of a light-emitting object, characterized by comprising:

( 8 ) having light receiving wavelength selection means for preferentially passing the reflected light of the laser slit light;
The object to emit light as described in ( 7 ) above, wherein when the image of the surface of the object irradiated with the laser slit light is picked up by the pair of stereo cameras, the light receiving wavelength selection means is interposed. 3D profile measuring device.
(9 ) A deviation amount between the first three-dimensional profile data and the second three-dimensional profile data is obtained by a predetermined method, and the obtained deviation amount is compared with a preset limit value. A deviation amount evaluation means for evaluating the deviation amount;
First data conversion means for translating the entire second three-dimensional profile data in the optical axis direction of the stereo camera pair;
A second data conversion means for rotating the entire second three-dimensional profile data around two axes orthogonal to the optical axis of the stereo camera pair and orthogonal to each other after the parallel movement ; Have
Using the first data conversion means and the second data conversion means, the second three-dimensional profile data is corrected to be corrected second three-dimensional profile data so that the deviation amount is minimized. ,
The modified second three-dimensional profile data is used instead of the original second three-dimensional profile data, and a union with the first three-dimensional profile data is calculated to obtain a three-dimensional profile of the entire object. The apparatus for measuring a three-dimensional profile of a light-emitting object according to ( 7) or (8) , wherein data is obtained.

(1 0 ) The position of each point of the first three-dimensional profile data and the second three-dimensional profile data is compared in the optical axis direction view of the stereo camera pair, and an overlapping point or a predetermined point Neighboring point pair setting means for setting a pair of points closer to the distance as a neighboring point pair;
Deviation distance deriving means for obtaining a deviation distance which is a distance in the optical axis direction of the neighboring point pair for at least four or more neighboring point pairs arbitrarily selected from the neighboring point group which is a set of the neighboring point pairs. And
The said deviation amount evaluation means calculates | requires each square sum of the said deviation distance as said deviation amount, The measuring apparatus of the three-dimensional profile of the light-emitting target object as described in said ( 9 ) characterized by the above-mentioned.

(1 1 ) The first data conversion means translates the entire second three-dimensional profile data in the optical axis direction of the stereo camera pair so that the deviation amount is minimized, As movement correction 3D profile data,
It said second data conversion means, After the translation, the translation modified three-dimensional profile data of the center of gravity, or coordinate the arithmetic mean of the coordinates of the values of each point of the translation modified three-dimensional profile data The parallel movement corrected three-dimensional profile data is rotated using the second data conversion means, with an arbitrary axis passing through the point having the value of and orthogonal to the optical axis of the stereo camera pair as a rotation axis. This is used as one-axis rotational movement corrected three-dimensional profile data, passes through the center of gravity of the one-axis rotational movement corrected three-dimensional profile data, and is orthogonal to the optical axis of the stereo camera pair and the arbitrary one axis. The rotational axis of the one-axis rotational movement correction 3D profile data is rotated again so that the amount of deviation is minimized, and the corrected second 3 Characterized by a source profile data, the (9) or three-dimensional profile of the measuring device of the object which emits according to (1 0).
(1 2 ) The stereo camera pair includes a first stereo camera pair that captures an image in the laser light imaging mode and a second stereo camera pair that captures an image in the ambient light imaging mode. The measuring device of the three-dimensional profile of the object to emit light according to any one of ( 7 ) to (1 1 ).

According to the present invention as described above (1), based on the first three-dimensional profile data obtained quantitatively by irradiating the laser slit light on the surface of the object, resulting in self-luminous and ambient light Correct the position of the second three-dimensional profile data was, by fusing the second three-dimensional profile data to the first three-dimensional profile data, the three-dimensional shape data with high precision and resolution Can be measured with high efficiency. Here, it originates in the difference in characteristics between imaging of laser slit light and imaging of self-light emission and ambient light. The shape data based on the laser slit light, which is the first three-dimensional profile data, has few errors and high accuracy, but has high directionality and has a low resolution because it is difficult to measure a particularly sharp shape change. On the other hand, the shape data (shade) by the self-light emission and the ambient light, which is the second three-dimensional profile data, has a higher resolution because the sharper shape change has clearer shadow information because the shape change is sharper. Since the change is steep, accuracy tends to decrease. Therefore, the second 3D profile data is merged with the first 3D profile data after the position of the second 3D profile data is changed on the basis of the highly accurate first 3D profile data. Thus, accuracy and resolution can be increased at the same time. Moreover, even if it is the elongate unevenness | corrugation extended in parallel with the laser slit light to irradiate, the unevenness | corrugation can be measured. Therefore, for example, even when high-temperature refractories inside the converter or inside the heating furnace are applied as objects, it is possible to easily estimate the presence or absence of refractories and the remaining thickness, and repair them without missing the machine. can do.

  Further, according to another feature of the present invention as described in (2) above, since the light receiving wavelength selection means for preferentially passing the reflected light of the laser slit light is used, the laser slit light is given priority over other light. It can be detected. Therefore, the output of the laser slit light can be further reduced, and the safety of measurement work can be obtained. Therefore, the use of a cover that prevents leakage of laser slit light and reflected light can be omitted.

According to another feature of the invention as in any of ( 3 ) to ( 5 ), the second three-dimensional profile is corrected after correcting the translational deviation of the second three-dimensional profile data. The deviation of rotational movement of data was reduced . Since the influence of the error due to the translational movement of the second three-dimensional profile data is larger than the influence due to the error of the rotational movement, a three-dimensional profile with higher accuracy and higher resolution can be obtained in this way.

If the apparatus of ( 7 ) is used, the effect by the method of (1) can be realized.
If the apparatus ( 8 ) is used, the effect of the method (2) can be realized.
If any one of the devices ( 9 ) to (1 1 ) is used, the effect of any one of the methods ( 3 ) to ( 5 ) can be realized.

(First embodiment)
DESCRIPTION OF EXEMPLARY EMBODIMENTS Hereinafter, a first embodiment of the invention will be described with reference to the drawings. The present embodiment is mainly an embodiment of the inventions (1), (2), (9), and (10).
1 to 3 are conceptual diagrams showing a stereo measurement apparatus according to an embodiment of the present invention. Specifically, FIG. 1 is a perspective view illustrating an example of the appearance of a stereo measurement device that obtains a pair of stereo images by laser slit light. FIG. 2 is a diagram illustrating an example of the concept of data flow and control in the stereo measurement apparatus. FIG. 3 shows an example of a measurement object, data obtained by irradiating the measurement object with laser light, and without irradiating the laser light (based on light by self-emission of the measurement object and ambient light). It is a perspective view which shows an example of the concept of fusion with data.

As shown in FIG. 1, in the stereo measurement device of the present embodiment, the laser light emitted from the laser light emitting means 1 is converted into laser slit light 3 by the slitting means 2 including, for example, a cylindrical lens, and the measurement target 4 A laser emission line 5 is generated on the surface. Then, the laser scanning means 6 scans the position of the laser emission line 5 in a direction substantially orthogonal to the major axis direction of the laser slit light 3. The stereo camera pairs 7a and 7b are arranged so that the scanning range of the laser bright lines 5 substantially coincides with the visual field range, and the laser bright lines 5 in the visual field are imaged by the stereo camera pairs 7a and 7b. The stereo camera pairs 7a and 7b are arranged so that the base line 8 of the stereo camera pair 7a and 7b (a line segment connecting the centers of the image pickup elements of the cameras) is substantially orthogonal to the laser emission line 5. The reflected light from the laser emission line 5 is adjusted by, for example, received light wavelength selection means 9a, 9b including a filter or the like and imaged by the stereo camera pair 7a, 7b.
As described above, in the present embodiment, the laser light emitting means 1, the slitting means 2, and the laser scanning means 6 realize a laser slit light irradiation means. Note that the slitting means 2 does not necessarily have to include a cylindrical lens. For example, a parabolic mirror may be provided instead of the cylindrical lens.

In addition, the stereo measurement apparatus of this embodiment includes a mode switching unit, a stereo image processing unit, a data fusion unit, and the like (not shown) in FIG. 1 (see FIG. 2).
The laser light emitting means 1, the stereo camera pair 7a, 7b, and the laser scanning means 6 are attached to the inside of a heat-resistant housing that is installed at a predetermined distance from the measurement object 4. This housing is not limited to the one shown in FIG. 1. For example, the region where the laser slit light 3 scanned by the laser scanning means 6 is emitted and the region where the reflected light from the laser emission line 5 is incident. The heat resistance of the device inside the housing may be further enhanced by closing the other portions.

Next, the relationship between the control in the stereo measurement apparatus and data such as image data will be described with reference to FIG.
In the mode switching unit 12, the imaging mode is switched between a laser beam imaging mode in which laser light is emitted from the laser emitting unit 1 and an ambient light imaging mode in which emission of laser light from the laser emitting unit 1 is stopped. The mode switching means 12 sends this switching signal to the laser light emitting means 1 and also sends a control signal necessary for this switching to the shutter means 13. That is, the mode switching unit 12 also controls the time (shutter opening time, ie, exposure time) during which the shutter unit 13 opens the shutters of the stereo camera pair 7a and 7b. Thereby, the measurement by the laser slit light 3 and the measurement by the light reflected from the ambient light 20 and the light (self-emission) 21 by the self-emission of the measurement object 4 can be switched. Note that the mode switching unit 12 may automatically switch between the laser light imaging mode and the ambient light imaging mode by computer control, or may be switched based on an operation by the user.

  In the measurement using the laser slit light 3, an image based on the reflected light of the plurality of laser emission lines 5 scanned by the laser scanning means 6 is interposed between the light receiving wavelength selection means 9 and the shutter means 13 and the stereo camera pairs 7 a and 7 b. Take an image with. Then, the stereo image processing means 10 performs stereo image processing (stereo viewing processing) on the image data output from the stereo camera pair 7a, 7b, and the first three-dimensional profile data 11 including a plurality of three-dimensional profile lines is obtained. obtain. In the laser beam imaging mode, the shutter opening time is adjusted to a short time during which reflected light can be imaged.

On the other hand, in the ambient light imaging mode, the shutter opening time (exposure time) is adjusted to a long opening time during which the weak ambient light 20 can be imaged without emitting the laser slit light 3. Then, the stereo image processing means 10 performs stereo image processing on the image data picked up by the stereo camera pair 7a, 7b through the light receiving wavelength selection means 9 and the shutter means 13, and consists of a plurality of three-dimensional profile lines. Second 3D profile data 16 is acquired.
As described above, in the stereo measurement device 100 of the present embodiment, two types of three-dimensional profile line groups (first three-dimensional profile data 11 are measured by measuring the same measurement object 4 in two imaging modes. And second 3D profile data 16). Then, the data fusion means 14 fuses two types of three-dimensional profile line groups to obtain three-dimensional profile data 17 with high accuracy and high resolution.

  In this embodiment, the stereo image processing means 10, the mode switching means 12, and the data fusion means 14 are, for example, an image input / output board for inputting image data captured by the stereo camera pair 7a, 7b, and This can be realized by using a personal computer provided with a communication interface for communicating with the laser emitting means 1 and the shutter means 13.

Next, referring to FIG. 3, the measurement object 4, the first three-dimensional profile data 11 of the measurement object 4 obtained in the laser light imaging mode, and the measurement obtained in the ambient light imaging mode An example of the concept of the second 3D profile data 16 of the object 4 and the 3D profile data 17 obtained by fusing the first 3D profile data 11 and the second 3D profile data 16 explain.
FIG. 3A is a perspective view showing an example of the measurement object 4, FIG. 3B is a perspective view showing an example of the first three-dimensional profile data 11, and FIG. FIG. 3 is a perspective view showing the second three-dimensional profile data 16, and FIG. 3D is a perspective view showing an example of the three-dimensional profile data 17. In the example shown in FIG. 3, the measurement object 4 includes a step 51 parallel to the laser slit light 3, a region (shadow) 52 that is darker than the overall average in self-emission, and a region (shadow) that is brighter than the average. 53, a concave region (step) 54 having an average brightness and a convex region (step) 55 are shown.

  The self-luminous light 21 of a high-temperature object changes in light emission according to the temperature, and tends to be light having a wavelength distribution according to Planck's formula. The ambient light 20 tends to have a wavelength distribution similar to that of the self light emission 21. Since the laser light has a single wavelength, the laser slit light 3 in the visible light region having a wavelength of 400 nm or more and 680 nm or less that is generally available can form the laser emission line 5 even at a low output.

First, in order to measure the shape of a refractory in a temperature range where light emission is weak at less than 1000 ° C., the laser slit light 3 in the visible light region having a constant wavelength of 400 nm to 680 nm (400 to 680 nm) can be used. .
On the other hand, since a higher-temperature object starts to emit self-luminous light 21 from red light, the laser slit light 3 having a shorter wavelength, that is, a hue such as green, blue, or purple, is used to increase the contrast of the laser emission line 5. It is desirable to use laser slit light 3 having a wavelength. For example, in order to measure the shape of a refractory having a temperature of 1000 ° C. or higher, it is desirable to use laser slit light 3 having a constant wavelength of 400 nm to 550 nm. However, since light having a wavelength of less than 400 nm becomes ultraviolet light, it is not visible light and may be harmful to the human body. Therefore, it is better not to use light having a wavelength of less than 400 nm as the laser slit light 3.
Furthermore, if the bright line, which is the reflected light of the laser slit light 3, is selectively imaged using the received light wavelength selection means 9, the contrast can be further increased, so that the measurement accuracy is improved.

  In general, since stereo image processing is measured and processed by a digital device, the measured image data becomes discrete points. Further, since the boundary where the luminance changes in the image is easily converted into data, the image data is likely to be a data point group arranged in a line when displayed as a drawing. Therefore, in the example shown in FIG. 3, the three-dimensional data point group that is the output of the stereo image processing means 10 is represented by a three-dimensional line segment that is formed by connecting these three-dimensional data point groups.

  As described above, FIG. 3B shows the first three-dimensional profile data (three-dimensional profile line group) 11 obtained from the reflected light of the laser emission line 5. As shown in FIG. 3B, in the first three-dimensional profile data 11, the continuous shape change is easily understood, but the step is easily lost. In other words, in the first three-dimensional profile data 11, it is easy to recognize a surface and it is difficult to recognize a ridge line or a valley line.

  FIG. 3C shows the second three-dimensional profile data (three-dimensional profile line group) 16 obtained from the shadow caused by the self-light emission 21 and the ambient light 20. As shown in FIG. 3C, in the second three-dimensional profile data 16, since a large change in shadow is easily recognized, the steps 54 and 55 can be easily obtained. The regions 52 and 53 where the shadow is generated are also regarded as features. In other words, in the second three-dimensional profile data 16, it is difficult to recognize a surface and it is easy to recognize a line and a luminance difference.

FIG. 3D shows the three-dimensional profile data 17 obtained by fusing two types of three-dimensional profile line groups (first three-dimensional profile data 11 and second three-dimensional profile data 16) by the data fusion means 14. It is a figure which shows an example.
As shown in FIG. 3D, for example, in the second three-dimensional profile data 16, the shape data of the regions 52 and 53 where the luminance changes are also obtained from the reflected light of the laser emission line 5. It can be seen that the vicinity of the regions 52 and 53 is flat when the three-dimensional profile data 11 is overlapped. Therefore, the three-dimensional profile data 17 obtained by fusing the first three-dimensional profile data 11 and the second three-dimensional profile data 16 can be effectively used to recognize the entire shape of the measurement object 4. Further, in the three-dimensional profile data 17 shown in FIG. 3D, the relationship between the step 51 and the entire shape becomes clear.

  As described above, in the present embodiment, the first three-dimensional profile data 11 obtained quantitatively by irradiating the surface of the measuring object 4 with the laser slit light 3, the self-luminous light 21, and the ambient light 20 are obtained. By fusing the second three-dimensional profile data 16, it is possible to measure three-dimensional shape data (three-dimensional profile data 17) with high accuracy and high resolution with high efficiency. Moreover, even if it is the elongate unevenness extended in parallel with the laser slit light 3 to be irradiated, the unevenness can be measured. Therefore, for example, even when the technique of the present embodiment is applied to a high-temperature refractory inside the converter or the inner wall of the heating furnace, the presence or absence of defects in the refractory and the remaining thickness can be easily estimated. Refractories can be repaired without missing.

  Further, since the light receiving wavelength selection means 9 that preferentially passes the reflected light of the laser slit light 3 is used, the laser slit light 3 can be detected with priority over other light. Therefore, the output of the laser slit light 3 can be further reduced, and the safety of measurement work can be obtained. Therefore, the use of a cover that prevents the leakage of the laser slit light 3 and the reflected light can be omitted.

(Second Embodiment)
Next, a second embodiment of the present invention will be described. This embodiment is different from the first embodiment described above mainly in the light emission form of laser slit light. Therefore, in the description of the present embodiment, the same parts as those of the first embodiment described above are denoted by the same reference numerals as those shown in FIGS. 1 to 3, for example, and detailed description thereof is omitted. The present embodiment is mainly an embodiment of the inventions (3) and (10).

FIG. 4 is a diagram illustrating an example of the concept of data flow and control in the stereo measurement apparatus 200.
In FIG. 4, when the mode switching unit 12 switches the imaging mode to the laser beam imaging mode, the laser slit light is intermittent pulsed light. Therefore, the mode switching unit 12 sends this switching signal to the pulse laser emitting unit 41. In addition, a control signal necessary for this switching is sent directly to the imaging synchronization means 15 without being sent directly to the shutter means 13.

The pulse laser emission means 41 emits intermittent pulse light. Specifically, the pulse laser emission means 41 has a characteristic of emitting intermittent pulse light having a light emission time of 1 / 100,000 or less, 1 / 10,000,000 or less (seconds) as the laser slit light 3. . The shutter means 13 has a characteristic that the exposure time of the stereo camera pair 7a, 7b can be adjusted to be longer than the light emission time of the pulse laser light emission means 41 and shorter than the light emission interval. Further, the imaging synchronization means 15 can cause the laser slit light to be emitted from the pulse laser emission means 41 in accordance with the opening time of the shutter by the shutter means 13 by the trigger signal from the mode switching means 12. The means 13 and the pulse laser emission means 41 are adjusted (controlled). In other words, the imaging synchronization means 15 is configured so that the exposure time of the stereo camera pair 7a, 7b includes the time during which the reflected light of the laser slit light is observed by the stereo camera pair 7a, 7b. Is synchronized with the laser slit light emission operation.
As described above, when the laser emission is a pulse, even if it is instantaneously output with high energy, it can be reduced to low energy on average. High energy can increase the S / N ratio, but increasing the output can lead to danger. Therefore, by making the laser emission pulse, the energy is high only at the moment of exposure, so the S / N ratio can be increased and averaged to be low energy.

  When the imaging mode is switched to the ambient light imaging mode, the mode switching unit 12 sends this switching signal to the pulse laser emission unit 41 and does not emit the laser slit light 3. In the ambient light imaging mode, the mode switching means 12 sends an appropriate control signal to the shutter means 13 without operating the imaging synchronization means 15, and the weak ambient light 20 can be imaged with the shutter opening time (exposure time). Adjust to a long shutter opening time (exposure time).

  With such a configuration, in the laser beam imaging mode of the present embodiment, imaging can be performed with a shorter exposure time than the configuration shown in FIG. 2 (the first embodiment), and the first obtained in the laser beam imaging mode. In the three-dimensional profile data 11, the influence of disturbance such as the self-light emission 21 can be further reduced. On the other hand, in the ambient light imaging mode in which the second three-dimensional profile data 16 obtained by the self-light emission 21 and the ambient light 20 is obtained, the same imaging as the configuration shown in FIG. 2 (first embodiment) is possible.

  As described above, in the present embodiment, since the laser slit light is intermittent pulse light, the exposure time of the pair of stereo cameras 7a and 7b at the time of imaging with the laser slit light is shortened. It is possible to reduce the influence of disturbance due to the self-light emission 21 emitted from the surface. That is, the S / N ratio of the stereo measurement apparatus 200 can be increased. Therefore, the output of the laser slit light can be further reduced. In addition, since the imaging of the laser slit light 3 and the imaging of the self-light emission 21 and the ambient light 20 can be switched only by adjusting the exposure time of the stereo camera pair 7a, 7b, both images (first three-dimensional profile data 11). And the second three-dimensional profile data 16) are almost free of optical differences, and the measurement accuracy can be further improved.

(Third embodiment)
Next, a third embodiment of the present invention will be described. In the present embodiment and the first and second embodiments described above, the fusion method of the first three-dimensional profile data 11 and the second three-dimensional profile data 16 is mainly different. Therefore, in the description of this embodiment, the same parts as those of the first and second embodiments described above are denoted by the same reference numerals as those shown in FIGS. . In addition, this embodiment is mainly embodiment of the invention of said (4)-(6), (11)-(13).

  FIG. 5 shows an example of the measurement object, the first three-dimensional profile data obtained by irradiating the measurement object with laser light, and the laser light without irradiation (light and ambient light by the self-emission of the measurement object). FIG. 6 shows an example of a concept of a method for fusing the obtained second three-dimensional profile data. FIG. 6 is a diagram illustrating an example of a concept following FIG. 5 for a method of fusing the first three-dimensional profile data and the second three-dimensional profile data.

  The first and second three-dimensional profile data 11 and 16 obtained by the two kinds of methods as shown in the first and second embodiments described above have no error. There is a difference in error characteristics between the first and second three-dimensional profile data 11 and 16. In the present embodiment, in consideration of this point, the first and second three-dimensional profile data 11 and 16 are merged to obtain three-dimensional profile data with higher accuracy and higher resolution.

5, FIG. 5 (a) is a perspective view showing an example of the measurement object 4, and FIG. 5 (b) is a projection view of the first three-dimensional profile data 11, which is a stereo camera pair 7a. 7b is a projection view of the optical axis direction view. FIG. 5C is a projection view of the second three-dimensional profile data 16, and is a projection view of the stereo camera pair 7a, 7b as viewed in the optical axis direction. FIG. 5D shows the first three-dimensional profile data 11 (FIG. 5B) and the second three-dimensional profile data 16 (FIG. 5C) as the optical axes of the stereo camera pairs 7a and 7b. It is the projection figure accumulated by the direction view.
As shown in FIG. 5 (d), the data fusion unit 14 compares the first and second measurement point groups constituting the first and second three-dimensional profile data 11 and 16, Then, a pair of measurement points closer than a predetermined distance determined in advance is obtained. In the following description, this pair of measurement points will be referred to as neighbor point pairs 56a, 56b, 56c, 56d. The two neighboring points constituting the neighboring point pair 56 are small, but have an interval (distance) in the optical axis direction of the stereo camera pair 7a, 7b. In the following description, this interval (distance) is referred to as a deviation distance. This deviation distance is graphically a distance in the optical axis direction between two neighboring points constituting the neighboring point pair 56. When the measurement points of the first and second three-dimensional profile data 11 and 16 are overlapped, the deviation distance is 0 (zero). As described above, in this embodiment, the data fusion unit 14 realizes a neighboring point pair setting unit.

FIG. 6 shows an example of a concept of a method for adjusting the three-dimensional profile data 17 obtained by superimposing the first and second three-dimensional profile data 11 and 16 as shown in FIG. FIG. For convenience of explanation, FIG. 6 shows only four neighboring point pairs 56a to 56d among the many neighboring point pairs 56.
FIG. 6A is a perspective view in which the first three-dimensional profile data 11 and the second three-dimensional profile data 16 are directly superimposed (that is, FIG. 6A is a perspective view of FIG. 5D). Is just different). In FIG. 6, both ends of the thick line segment are neighboring point pairs 56, and the lengths indicate deviation distances 57.
The data fusion means 14 appropriately distributes the numerical value of the deviation distance 57 in the neighboring point pair group that is a set of the neighboring point pairs 56, and includes at least four neighboring point pairs 56 from the neighboring point group. If possible, about 10 are selected arbitrarily, and the deviation distance 57 (deviation distances 57a to 57d in the example shown in FIG. 6A) in the selected neighboring point pair 56 is obtained, and the sum of squares of each deviation distance 57 (square). Is calculated as the amount of deviation. As described above, in the present embodiment, the data fusion unit 14 realizes the deviation distance deriving unit and the deviation amount evaluating unit.

When the entire three-dimensional profile (second three-dimensional profile data 16) obtained by the ambient light 20 and the self-light emission 21 is moved in the optical axis direction of the stereo camera pair 7a, 7b, the deviation amount (deviation distance 57) Increases or decreases. Therefore, the data fusion means 14 has a deviation distance 57 when the second three-dimensional profile data 16 is moved in the optical axis direction of the stereo camera pair 7a, 7b within a predetermined range on the camera side and the opposite camera side. The movement distance of the second three-dimensional profile data 16 that minimizes is obtained. Then, the data fusion means 14 translates the entire second three-dimensional profile data 16 in the direction of the optical axis of the stereo camera pair 7a, 7b by a movement distance that minimizes the deviation distance 57. In the following description, the second three-dimensional profile data 16 moved by the movement distance that minimizes the deviation distance 57 is referred to as parallel movement corrected three-dimensional profile data. FIG. 6B shows the first three-dimensional profile data 11 and this parallel movement corrected three-dimensional profile data superimposed on each other.
As described above, in the present embodiment, the first data conversion unit is realized by the data fusion unit 14.

Next, the data fusion means 14 uses the arithmetic mean point having the arithmetic mean value of each coordinate value of each point of the translation corrected three-dimensional profile data as the coordinate value, or the centroid of the translation corrected three-dimensional profile data as the coordinate value. Find the barycentric point you have.
The translation-corrected three-dimensional profile data is rotated about the horizontal axis 58 (around the horizontal axis 58) passing through the arithmetic mean point or the center of gravity 59 and orthogonal to the optical axis of the stereo camera pair 7a, 7b. When moved, the amount of deviation increases or decreases. Therefore, the data fusion means 14 sets the parallel movement corrected three-dimensional profile data in the positive and negative directions (in the direction of the horizontal axis 58) in the positive and negative directions with the horizontal axis 58 as the rotation axis so that the deviation amount is minimized. Rotate within range. In the following description, the translation corrected three-dimensional profile data rotated in this way is referred to as uniaxial rotational movement corrected three-dimensional profile data. FIG. 6C is a perspective view showing the first three-dimensional profile data 11 and the one-axis rotational movement corrected three-dimensional profile data in an overlapping manner.

The uniaxial rotation movement (around the vertical axis 60) is performed with the vertical axis 60 passing through the center of gravity or arithmetic mean point 59 and orthogonal to both the optical axis and the horizontal axis 58 of the stereo camera pair 7a, 7b. When the corrected three-dimensional profile data is rotationally moved, the amount of deviation increases or decreases. Therefore, the data fusion means 14 determines the uniaxial rotational movement corrected three-dimensional profile data in both positive and negative directions (in the direction of the vertical axis 60) with the vertical axis 60 as the rotation axis so that the deviation amount is minimized. Rotate within the range. In the following description, the one-axis rotational movement corrected three-dimensional profile data thus rotated is referred to as corrected second three-dimensional profile data. FIG. 6D is a perspective view showing the first three-dimensional profile data 11 and the modified second three-dimensional profile data in an overlapping manner.
As described above, in the present embodiment, the second data conversion unit is realized by the data fusion unit 14.

Then, the data fusion means 14 calculates the union of the first three-dimensional profile data 11 and the modified second three-dimensional profile data, and the three-dimensional profile of the entire surface of the measurement object 4 with the calculated union. Data.
The movement of the second three-dimensional profile data 16 performed as described above is preferably performed when the deviation amount (deviation distance 57) is larger than a preset limit value.

The first three-dimensional profile data 11 has higher positional accuracy than the second three-dimensional profile data 16. On the other hand, the second three-dimensional profile data 16 has a higher profile shape accuracy than the first three-dimensional profile data 11. Therefore, in the present embodiment, by correcting only the position of the second three-dimensional profile data 16 on the basis of the first three-dimensional profile data 11, and then fusing with the first three-dimensional profile data 11, High-precision and high-resolution three-dimensional profile data can be obtained.
Further, the influence of the translational movement of the second three-dimensional profile data 16 is larger than the influence of the rotational movement error. Therefore, when the deviation of the rotational movement of the second three-dimensional profile data 16 is reduced after the translational deviation of the second three-dimensional profile data 16 is corrected as in the present embodiment, the resolution is higher. A high three-dimensional profile can be obtained.

  As described above, in the present embodiment, since the position of the second three-dimensional profile data 16 is corrected and merged with the first three-dimensional profile data 11, the shape data of the measurement object 4 can be obtained with high accuracy. it can. This is due to the difference in characteristics between the imaging of the laser slit light 3 and the imaging of the self-light emission 21 and the ambient light 20. The first three-dimensional profile data 11, which is shape data by the laser slit light 3, has small errors and high accuracy, but has strong directivity and has a low resolution because it is difficult to measure a particularly steep shape change. On the other hand, the second three-dimensional profile data, which is the shape data by the self-light-emitting 21 and the ambient light 20, has a higher resolution because it has clearer shadow information as the shape change is less directional and sharp, but the original shape change. The accuracy is likely to decrease due to the steepness. Therefore, by combining the second three-dimensional profile data 16 with the high-precision first three-dimensional profile data 11 as a reference, the accuracy and the resolution can be increased simultaneously.

  Note that a single pair of stereo cameras is sufficient to achieve the effects of the above-described embodiments. However, the number of stereo camera pairs may not be one, and a plurality of stereo camera pairs may be provided according to the imaging mode. For example, a stereo camera pair (first stereo camera pair) for obtaining the first three-dimensional profile data 11 obtained in the laser light imaging mode and the second three-dimensional profile data 16 obtained in the ambient light imaging mode are obtained. A stereo camera pair to be obtained (second stereo camera pair) may be separately installed, and two (2 pairs) of stereo camera pairs may be provided.

  In the repair work of the refractory of the molten steel pan, when the refractory was updated, the three-dimensional profile data of the inner wall surface of the pan, that is, the inner surface of the iron skin was measured by the apparatus (method) of the present invention. Thereafter, in a state where the refractory was applied and operated, the three-dimensional profile data of the inner wall surface of the pan was measured again by the method of the present invention. At this time, the measurement of the inner surface of the iron skin before the refractory renewal and the measurement of the inner wall surface of the pan after the construction of the refractory, the position of the molten steel pan and the relative position of the installed device of the present invention are It has been adjusted to be the same. By applying the apparatus (method) of the present invention even on a surface having a small optical characteristic such as a refractory immediately after construction, or the like, the presence or absence of a defect in the refractory and the remaining thickness could be measured accurately.

  And the difference of these two three-dimensional profile data becomes the thickness of the constructed refractory. The thickness of the refractory was an average of 150 mm, a minimum of 120 mm, and a maximum of 160 mm. After this molten steel pan was operated for 100 ch by proper production, the three-dimensional profile data of the refractory on the inner surface was measured with the apparatus (method) of the present invention between the operations every 10 ch. Based on the conventional knowledge, the criterion is to stop proper production and reconstruct the refractory when the minimum remaining thickness of the refractory reaches 40 mm.

  When the measurement by the apparatus (method) of the present invention was performed after 250 ch operation, the minimum remaining thickness of the refractory was 50 mm. Therefore, the reduction rate of the minimum remaining refractory was estimated to be 5 mm every 10 ch. The maximum life expectancy of refractory was considered to be 270ch. However, in the measurement performed after 260 ch operation, it was found that the refractory had groove-like wear, and the minimum remaining thickness of the refractory in the region where the wear occurred was 30 mm, and the reference 40 mm Below. For this reason, production was discontinued. Due to the shape of the groove-like wear, the wear rate of the refractory was about twice the average wear rate. If the production was continued up to 270 ch as originally planned, the molten steel was eroded into the iron skin, and there was a possibility that production loss due to molten steel spilling due to the molten iron core and the safety of the work environment would be reduced. With the apparatus (method) of the present invention, the remaining thickness and wear of the refractory can be measured appropriately with high accuracy and in detail, so that it is possible to avoid production loss and a decrease in the safety of the work environment.

Similarly, by using the apparatus (method) of the present invention, the use in proper production can be continued until the minimum remaining thickness of the refractory is as close as possible to the standard value of 40 mm. Opportunities for production interruptions were minimized, production costs were reduced, and production volumes were expanded.
In the examples, it was used to measure the wear of the refractory in the converter, but the present invention is not limited to this, and any measuring object that self-lumines at high temperatures can be used. The shape of such a measurement object can be measured. For example, the present invention may be used to measure consumption due to oxidation of a refractory on the lining of a heating furnace or a slab heated in the heating furnace.

  It should be noted that each of the above-described embodiments and examples is merely an example of implementation in carrying out the present invention, and the technical scope of the present invention should not be construed in a limited manner. Is. That is, the present invention can be implemented in various forms without departing from the technical idea or the main features thereof.

It is a perspective view which shows the 1st Embodiment of this invention and shows an example of the external appearance of a stereo measurement apparatus, and is a figure explaining the state which obtains a pair of stereo image by a laser slit light. It is a figure which shows the 1st Embodiment of this invention and shows an example of the concept of the flow of data and control in a stereo measuring device. The 1st Embodiment of this invention is shown, An example of a measurement object, and an example of the concept of fusion of the data obtained by irradiating a measurement object with a laser beam, and the data obtained without irradiating a laser beam, FIG. It is a figure which shows the 2nd Embodiment of this invention and shows an example of the concept of the flow of data and control in a stereo measuring device. The 3rd Embodiment of this invention is shown, An example of a measurement object, 1st three-dimensional profile data obtained by irradiating a measurement object with a laser beam, and the 2nd obtained without irradiating a laser beam It is a figure which shows an example of the concept of the method which unites 3D profile data. It is a figure which shows the example of the concept following FIG. 5 of the method which shows the 3rd Embodiment of this invention and fuse | melts 1st three-dimensional profile data and 2nd three-dimensional profile data.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Laser light emission means 2 Slit-izing means 3 Laser slit light 4 Measurement object 5 Laser bright line 6 Laser scanning means 7a, 7b Stereo camera pair 8 Base line (line segment which connects the center of the image sensor of each camera)
9a, 9b Light receiving wavelength selection means 10 Stereo image processing means 11 First three-dimensional profile data 12 Mode switching means 13 Shutter means 14 Data fusion means 15 Imaging synchronization means 16 Second three-dimensional profile data 17 Three-dimensional profile data 20 Environment Light 21 Self-emission 41 Pulse laser emission means 51 Parallel step 52 Area darker than average (shadow)
53 Area brighter than average (shade)
54 Recessed area (step)
55 Convex area (step)
56a, 56b, 56c, 56d Neighboring point pair 57 Deviation distance 58 Horizontal axis 59 Arithmetic mean point or barycentric point 60 Vertical axis

Claims (12)

Stereo measurement using a pair of stereo cameras installed at a predetermined interval to obtain a three-dimensional profile by imaging a light emitting object, and irradiating the object with laser slit light having a constant wavelength between 400 and 680 nm. In the measurement method for measuring the three-dimensional profile of a light emitting object,
The imaging mode is switched to an imaging mode (hereinafter referred to as laser beam imaging mode) in which the laser slit light is irradiated by mode switching means,
After the imaging mode is switched to the laser light imaging mode, irradiate the laser slit light on the surface of the object,
Taking an image of the surface of the object irradiated with the laser slit light by the stereo camera pair,
By performing stereo image processing on the captured image, first three-dimensional profile data based on the laser slit light is obtained,
The imaging mode is switched to an imaging mode (hereinafter referred to as ambient light imaging mode) in which the laser slit light is not irradiated by the mode switching means,
After the imaging mode is switched to the ambient light imaging mode, a shadow generated on the surface of the object is generated by the stereo camera pair based on the ambient light and the self-light emitted from the surface of the object. Take an image containing
Stereo image processing is performed on the captured image to obtain second three-dimensional profile data based on the ambient light and the self-light emission,
The position of the second three-dimensional profile data is corrected with reference to the first three-dimensional profile data, and then the second three-dimensional profile data is merged with the first three-dimensional profile data. The three-dimensional profile data of the whole object is obtained by the above method, and the method for measuring the three-dimensional profile of the object to emit light is provided.   When obtaining the first three-dimensional profile data by the stereo image processing, the light receiving wavelength selection means that preferentially allows the reflected light of the laser slit light to pass through the object irradiated with the laser slit light. The method for measuring a three-dimensional profile of a light-emitting object according to claim 1, wherein an image of a surface is taken. In the first three-dimensional profile data and the second three-dimensional profile data, a deviation amount between the two data is evaluated by a predetermined method,
If pre said deviation amount from the limits set in large, the entire second three-dimensional profile data after moving parallel to the optical axis of the stereo camera pair, perpendicular to the optical axis of the stereo camera pair and performing that you rotate moving the entire second three-dimensional profile data as the rotation axis of two mutually orthogonal axes, modifying said second three-dimensional profile data so that the deviation amount becomes minimum As a modified second 3D profile data,
The modified second three-dimensional profile data is used in place of the original second three-dimensional profile data and fused with the first three-dimensional profile data to obtain three-dimensional profile data of the entire object. wherein the three-dimensional profile of the measurement method of the object which emits according to claim 1 or 2. Compare the positions of the points of the first three-dimensional profile data and the second three-dimensional profile data when viewed in the optical axis direction of the stereo camera pair, and the overlapping points or points closer than a predetermined distance A pair of neighbors,
The distance in the optical axis direction of the neighboring point pair as a deviation distance,
Obtaining the deviation distance for at least four or more neighboring point pairs arbitrarily selected from among neighboring point groups that are a set of neighboring point pairs in the entire first and second three-dimensional profile data;
4. The method for measuring a three-dimensional profile of a light-emitting object according to claim 3 , wherein a sum of squares of the obtained deviation distances is obtained as the deviation amount. Wherein such deviation amount becomes minimum, after the second three-dimensional profile data across moved parallel to the optical axis of the stereo camera pair, which translation modifications three-dimensional profile data,
It passes through the center of gravity of the translation correction three-dimensional profile data or the point having the arithmetic mean of the coordinate values of each point of the translation correction three-dimensional profile data as the coordinate value, and on the optical axis of the stereo camera pair Rotating and moving the translation correction three-dimensional profile data with an arbitrary one axis orthogonal to the rotation axis, this as one axis rotation movement correction three-dimensional profile data,
The one-axis rotational movement correction three-dimensional profile data with the axis passing through the center of gravity of the one-axis rotational movement correction three-dimensional profile data and orthogonal to the optical axis of the stereo camera pair and the arbitrary one axis as the rotation axis The three-dimensional object to be lit according to claim 3 or 4 , wherein the three-dimensional profile data is rotated again so that the amount of deviation is minimized, and this is used as the modified second three-dimensional profile data. Profile measurement method. The stereo camera pair includes a first stereo camera pair that captures an image in the laser light imaging mode and a second stereo camera pair that captures an image in the ambient light imaging mode. Item 6. The method for measuring a three-dimensional profile of a light-emitting object according to any one of Items 1 to 5 . In a three-dimensional profile measuring apparatus that irradiates an object to be emitted with laser slit light having a constant wavelength of 400 to 680 nm and performs stereo measurement using a pair of stereo cameras installed at predetermined intervals.
Mode switching means for switching an imaging mode between an imaging mode for irradiating the laser slit light (hereinafter referred to as a laser light imaging mode) and an imaging mode for not irradiating the laser slit light (hereinafter referred to as an ambient light imaging mode);
Laser slit light irradiation means for irradiating the surface of the object with the laser slit light;
The image of the surface of the object irradiated with the laser slit light, and the image including the shadow generated on the surface of the object based on the ambient light and the self-light emitted from the surface of the object. A pair of stereo cameras for imaging
The image of the surface of the object irradiated with the laser slit light is image-processed to obtain first three-dimensional profile data, and a shadow generated on the surface of the object based on the ambient light and the self-light emission. Stereo image processing means for obtaining second three-dimensional profile data by image processing an image including:
The position of the second three-dimensional profile data is corrected with reference to the first three-dimensional profile data, and then the second three-dimensional profile data is merged with the first three-dimensional profile data. By means of data fusion means for obtaining three-dimensional profile data of the entire object,
An apparatus for measuring a three-dimensional profile of an object that emits light. Receiving wavelength selection means for preferentially passing the reflected light of the laser slit light,
The object for emitting light according to claim 7 , wherein when the image of the surface of the object irradiated with the laser slit light is picked up by the pair of stereo cameras, the light receiving wavelength selection means is interposed. 3D profile measuring device. The amount of deviation between the first three-dimensional profile data and the second three-dimensional profile data is obtained by a predetermined method, and the amount of deviation is compared with a predetermined limit value. A deviation amount evaluation means for evaluating
First data conversion means for translating the entire second three-dimensional profile data in the optical axis direction of the stereo camera pair;
A second data conversion means for rotating the entire second three-dimensional profile data around two axes orthogonal to the optical axis of the stereo camera pair and orthogonal to each other after the parallel movement ; Have
Using the first data conversion means and the second data conversion means, the second three-dimensional profile data is corrected to be corrected second three-dimensional profile data so that the deviation amount is minimized. ,
The modified second three-dimensional profile data is used instead of the original second three-dimensional profile data, and a union with the first three-dimensional profile data is calculated to obtain a three-dimensional profile of the entire object. The apparatus for measuring a three-dimensional profile of a light-emitting object according to claim 7 or 8 , wherein data is obtained. Compare the position of each point of the first three-dimensional profile data and the second three-dimensional profile data in the optical axis direction view of the stereo camera pair, and the overlapping point or a point closer than a predetermined distance Neighboring point pair setting means for setting a pair of as neighboring point pairs;
Deviation distance deriving means for obtaining a deviation distance which is a distance in the optical axis direction of the neighboring point pair for at least four or more neighboring point pairs arbitrarily selected from the neighboring point group which is a set of the neighboring point pairs. And
10. The apparatus for measuring a three-dimensional profile of a light-emitting object according to claim 9 , wherein the deviation amount evaluation unit obtains each square sum of the deviation distances as the deviation amount. The first data conversion means translates the entire second three-dimensional profile data in the direction of the optical axis of the stereo camera pair so as to minimize the deviation amount, and translates the same into the translation-corrected three-dimensional Profile data,
It said second data conversion means, After the translation, the translation modified three-dimensional profile data of the center of gravity, or coordinate the arithmetic mean of the coordinates of the values of each point of the translation modified three-dimensional profile data The parallel movement corrected three-dimensional profile data is rotated using the second data conversion means, with an arbitrary axis passing through the point having the value of and orthogonal to the optical axis of the stereo camera pair as a rotation axis. This is used as one-axis rotational movement corrected three-dimensional profile data, passes through the center of gravity of the one-axis rotational movement corrected three-dimensional profile data, and is orthogonal to the optical axis of the stereo camera pair and the arbitrary one axis. The rotational axis of the one-axis rotational movement correction 3D profile data is rotated again so that the amount of deviation is minimized, and the corrected second 3 Characterized by a source profile data, emitting three-dimensional profile of the measuring device of the object to be according to claim 9 or 1 0. The stereo camera pair includes a first stereo camera pair that captures an image in the laser light imaging mode and a second stereo camera pair that captures an image in the ambient light imaging mode. emitting three-dimensional profile of the measuring device of the object to be according to any one of claim 7-1 1.

Images