KR20120051875A - Apparatus and method for monitoring a structure using motion capture method - Google Patents

Abstract

PURPOSE: A structure monitoring apparatus and a method using capture of motions are provided to measure an accurate deformation rate by measuring an absolute strain value of a structure by analyzing images photographed a position of a market. CONSTITUTION: A structure monitoring apparatus using capture of motions comprises a marker unit(100), a camera unit(200), an image processing unit(300), a coordinate conversion unit(400), and a deformation measuring unit(500). The market comprises a plurality of markets installed in a structure. The image processing unit filters the marker images photographed by a plurality of cameras. The coordinate conversion unit calculates three-dimensional coordinates by using the filtered marker images. The deformation measuring unit measures a deformation rate of the structure by analyzing a value of the three-dimensional coordinates calculated by the coordinate conversion unit.

Description

Structure monitoring apparatus and method using motion capture {APPARATUS AND METHOD FOR MONITORING A STRUCTURE USING MOTION CAPTURE METHOD}

The present invention relates to a monitoring apparatus and method for measuring strain on a structure. In particular, the present invention relates to a monitoring apparatus and method for measuring strain of a structure using optical motion capture.

Structural monitoring is to periodically assess the safety and usability of the structure to signal damage in advance by unexpected loads and to perform reinforcement. The safety evaluation of the structure is evaluated by stresses such as axial force and moment acting on the member, and the usability evaluation of the structure is evaluated by deflection of the member, inter-layer displacement of the building, and vibration.

It is rare to assess either safety or usability in monitoring structures. In order to accurately predict the structure response through the monitoring of the structure, the structure shape and system of the structure should be identified. However, due to construction accuracy limitations, there is always an error from the drawing, and no structure is built 100% accurately with the designed drawing. Also, as the properties of the material change over time, the response to the load on the structure can vary. Therefore, it is not possible to grasp the behavior of the structure by using only one monitoring measuring device used for safety or usability evaluation, and comprehensive monitoring to grasp the characteristics and behavior of the material of the changing structure by evaluating safety and usability together. You need a system.

Monitoring of structures is largely divided into monitoring at the member level and monitoring at the structure level. The monitoring of the member level measures the strain and displacement to measure the magnitude and stress of the load on the member. Sensors used to measure strain include an electrical resistive sensor and an optical strain sensor, but these strain sensors have a problem in that they can only evaluate stress in the local region of the member. In order to overcome this problem, a measurement technique based on average strain has been developed, but the deformation shape over the member length is not known.

Sensors used for monitoring at the structure level include accelerometers, GPS, LIDAR (Light Detection And Ranging), and camera-based vision-based measurement techniques. Accelerometers are mainly installed on the top floors of buildings and are used for evaluating the usability of buildings, and very accurate measurements are possible, but errors due to numerical integration occur and measurement of static displacements is difficult to measure the displacement of structures. .

GPS can measure two-dimensional absolute coordinates, but it can't measure smoothly when there are high structures around it, so it can only measure top-level displacement. It can also measure up to 25Hz as of 2010, but it lacks the ability to analyze dynamic components such as acceleration.

LIDAR has the advantage of being able to measure three-dimensional information relatively easily and remotely, but it is impossible to understand the dynamic displacement, acceleration, vibration, etc. of the structure. Because there is no method yet, there are limitations to the measurement.

One of the most used sensors for measuring displacement of structures is the Linear Variable Differential Transformer (LVDT). LVDTs differ in the limits and precision of displacements that can be measured depending on the type, but they have the advantage of being easy to use because the results are reliable and the installation and principle are simple. However, in order to measure the structure with LVDT, there is a problem that it is difficult to apply to large structure because reference points for fixing the sensor should be near the structure.

As such, existing monitoring techniques can be usefully used to measure stresses and deformations on structures. However, strain-based metrology at the member level relies on relative strain values, so the exact stress is unknown. This is because it is a measure of how much stress is reduced or increased in the initial state while assuming that the stress at the time of the first sensor installation is zero.

In addition, accelerometers are difficult to measure the deformed shape of the structure, it is difficult to apply to the measurement of atypical structures because only data about one axis, such as strain sensors can be measured. In fact, the higher the height of the building, the greater the torsion of the building due to wind load, which has a problem that it is almost impossible to accurately measure the torsion angle using a sensor based on one axis.

Structure monitoring apparatus and method using motion capture according to the present invention aims to solve the following problems.

First, the structural monitoring method based on image processing is to measure the absolute deformation value rather than the relative deformation value.

Second, the strain measurement of the structure is to be performed in real time using an image processing technique.

Third, three-dimensional deformation of the structure is measured by using a motion capture technique.

The solution to the problem of the present invention is not limited to those mentioned above, and other solutions not mentioned can be clearly understood by those skilled in the art from the following description.

Structure monitoring apparatus using a motion capture according to the present invention is a marker unit including a plurality of markers (markers) installed in the structure, a camera unit including a plurality of cameras for photographing the markers, filtering the marker image taken by the plurality of cameras An image processing unit, a coordinate measuring unit calculating a 3D coordinate of the marker using the marker image filtered by the image processing unit, and a strain measuring unit measuring the strain of the structure by analyzing the 3D coordinate value calculated by the coordinate converting unit. do.

The marker of the marker unit according to the present invention is characterized in that it is installed on the structure according to the size, shape, predetermined deformation characteristics of the structure.

The marker according to the present invention is an active marker or a passive marker, and in the case of the passive marker, an apparatus for generating infrared rays or light is disposed around the marker installed in the structure.

The image processing unit according to the present invention is characterized by performing filtering using a Generalized-Cross-Validatory Spline Smoother (GCVSPS) algorithm.

The coordinate transformation unit according to the present invention is characterized in that it comprises a calibration unit for setting the initial reference for the three-dimensional coordinates and a coordinate calculation unit for calculating the three-dimensional coordinates for the marker image based on the initial virtual three-dimensional coordinates.

The calibration unit according to the present invention sets the zero coordinates (0,0,0) as a reference for the three-dimensional coordinates, calculates the distance and direction of the camera and zero coordinates, and sets the relationship between the camera and the coordinates, The coordinate correction value may be determined in advance according to the type and performance.

The calibration unit according to the present invention is characterized in that when the position of one or more cameras is changed, resetting the initial reference for the three-dimensional coordinates.

If the calibration unit according to the present invention has one or more cameras whose positions have not changed when the positions of one or more cameras are changed, the calibration unit may determine the relationship between the coordinates of the cameras whose positions have not been changed and the coordinates. It is characterized by resetting the initial reference for the three-dimensional coordinates by the operation.

The coordinate calculation unit according to the present invention is characterized by calculating a three-dimensional coordinates for the marker using a motion capture algorithm.

The structure monitoring method using motion capture according to the present invention includes a step S1 of photographing a plurality of markers installed on a structure with a plurality of cameras and setting an initial reference of three-dimensional coordinates for the structure.

Structure monitoring method using a motion capture according to the present invention includes the step S2 of filtering the image taken by a plurality of cameras a plurality of markers installed on the structure.

The structure monitoring method using the motion capture according to the present invention includes a step S3 of using the marker image filtered in step S2 to calculate three-dimensional coordinates of the marker.

The structure monitoring method using the motion capture according to the present invention includes a step S4 in which the strain of the structure is measured by analyzing the three-dimensional coordinates calculated in step S3.

The marker according to the invention is characterized in that it is installed on the structure according to the size, shape, predetermined deformation characteristics of the structure.

The marker according to the present invention is an active marker or a passive marker, and in the case of the passive marker, an apparatus for generating infrared rays or light is disposed around the marker installed in the structure.

Step S2 according to the present invention is characterized in that it is filtered using the Generalized-Cross-Validatory Spline Smoother (GCVSPS) algorithm.

The step S1 according to the present invention includes a step S1-1 in which zero coordinates (0,0,0), which are a reference to three-dimensional coordinates, in the image photographed by the camera are set.

Step S1 according to the present invention includes step S1-2 in which the relationship between the camera and the coordinates is established by calculating the distance and direction of the camera and the zero coordinate.

Step S1 according to the present invention includes step S1-3 in which coordinate correction values are determined according to the type and performance of the camera.

In the step S1 according to the present invention, when the position of one or more cameras among the cameras is changed, the initial reference for the 3D coordinates is reset.

In the step S1 according to the present invention, if there is at least one camera whose position has not changed when the position of one or more cameras among the cameras is changed, the position between the changed camera and the coordinates based on the relationship between the camera whose position has not changed and the coordinates. The initial criterion for the three-dimensional coordinates is reset by calculating the relation of.

Step S3 according to the present invention is characterized in that the three-dimensional coordinates for the marker are calculated using a motion capture algorithm.

Step S4 according to the present invention is characterized in that the three-dimensional strain, deformation acceleration and torsion of the structure are measured by analyzing the three-dimensional coordinates calculated in step S3.

Structure monitoring apparatus and method using motion capture according to the present invention has the following effects.

First, since the absolute deformation value of the structure is measured by analyzing the image photographing the position of the marker with the camera, accurate strain measurement is possible.

Second, accurate and rapid strain measurement can be performed using image information captured in real time.

Third, since three-dimensional deformation can be measured by using a motion capture technique, it is possible to measure three-dimensional strain, acceleration of deformation, and distortion of a structure, unlike conventional methods.

The effects of the present invention are not limited to those mentioned above, and other effects not mentioned can be clearly understood by those skilled in the art from the following description.

1 is a block diagram schematically showing the configuration of a structure monitoring apparatus using motion capture according to the present invention.
2 is a block diagram showing a basic system for motion capture according to the present invention.
3 shows the principle of determining the position of the marker by reflecting infrared rays to the marker.
4 (a) and 4 (b) are graphs showing the results before and after the marker image filtering according to the present invention, respectively.
5 is a coordinate diagram for explaining a process of calculating three-dimensional coordinates according to the present invention.
6 is a flowchart illustrating a procedure of a structure monitoring method using motion capture according to the present invention.
7 is a flowchart specifically showing step S1 of the present invention.

Hereinafter, an apparatus and method for monitoring a structure using motion capture will be described in detail with reference to the accompanying drawings.

First, the motion capture technique, which is the core of the present invention, will be briefly described. There are three types of motion switchers: mechanical, magnetic and optical. Although mechanical data can obtain accurate data by calculating angle movements in human joints, etc., there are disadvantages in that the sensor is mounted directly on the object to be measured, the number of sensors is limited, and there is a limit to the behavior that can be measured. . Magnetic method is to install the magnetic field generating device and measure the displacement in space based on the change of the magnetic field measured by each sensor when the measuring object moves near. Compared to optical motion capture, it is relatively inexpensive and enables real-time processing of data.

Optical is a very accurate method of motion capture. Optical motion capture is a method in which two or more cameras measure markers in two-dimensional coordinates to calculate three-dimensional spatial coordinates. It can shoot at high speed, so there is almost no data lost and there are no restrictions on what can be measured. In addition, if the area to be measured is in the camera's field of view, it is more efficient to measure a large number of areas because only a marker can be measured without adding a camera. The present invention is based on the use of an optical motion capture technique.

1 is a block diagram schematically showing the configuration of a structure monitoring apparatus using motion capture according to the present invention. 2 is a block diagram showing a basic system for motion capture according to the present invention.

Structure monitoring apparatus using a motion capture according to the present invention is a marker unit 100 including a plurality of markers (markers) installed in the structure, a camera unit 200 including a plurality of cameras to shoot the marker, a plurality of cameras The image processing unit 300 to filter the photographed marker image, the coordinate transformation unit 400 and the coordinate transformation unit 400 calculated by the three-dimensional coordinates for the marker using the marker image filtered by the image processing unit Strain measurement unit 500 for measuring the strain of the structure by analyzing the coordinate value.

The present invention is to monitor the structure, the structure includes a variety of civil engineering structures, such as bridges, dams, plants, as well as building structures such as skyscrapers.

The marker of the marker unit 100 is preferably installed in the structure according to the size, shape, predetermined deformation characteristics of the structure. The location where the marker is to be installed should be selected for the best position to measure the strain of the structure. This can be determined by the size, shape of the structure, the geological conditions in which it is located, and the climatic conditions in which it is located.

Furthermore, if there is a site that needs to be focused on a specific structure, a marker must be installed at that site. The above-described deformation characteristics mean various factors that may affect the deformation rate, such as areas to be observed in accordance with the structure, geological location of the structure, and climatic conditions of the area in which the structure is located.

Optical motion capture can be divided into two types depending on how the camera recognizes markers. First, there is a passive marker method that tracks the position of the marker by reflecting infrared rays or light to the marker by adding a camera or other equipment. The other is the light emitted directly from the marker so that the camera can determine the position of the marker. There is an active marker method for tracking. There are many ways to capture optical motion, such as using passive, active markers, or using light or infrared light.

3 illustrates a principle of determining the position of the marker by reflecting infrared rays to the marker and returning the infrared rays back to the camera. When infrared rays are reflected on the marker, the captured image has different accuracy depending on the resolution of the camera.

The infrared light is emitted through the infrared generating device, and the emitted infrared light is reflected by the marker. At this time, when the infrared generating device is installed in various directions to emit infrared rays, the center of the marker is brightest, and the camera finds the center of the marker by finding the position of the brightest shining pixel. The theoretical number of cameras required to obtain three-dimensional coordinates is two, but the accuracy of the camera increases as the number of cameras increases.

On the other hand, since the structure is usually large in size, it is preferable to photograph with a camera from a long distance. This is because the number of cameras can be reduced if the entire structure can be photographed with a good camera.

Various factors can be considered for the distance shooting. One of them is infrared light or light emitted to detect the marker.In order to get farther away from the infrared light, install an infrared generator around the marker installed on the structure rather than emitting and reflecting infrared light from the camera. Thus, it is preferable to allow infrared rays or the like to proceed for a longer distance. It is preferable to provide a plurality of infrared ray generating devices so that infrared rays are emitted from various directions.

In another embodiment, the marker itself installed on the structure may be enlarged or the marker may be photographed for a long distance by increasing the brightness of the marker.

Currently commercially available LIDAR is a laser measurement technique that can be up to 1 km away. Thus, as another example, if the laser is emitted from the marker and the coordinates are acquired from the camera, long distance measurements will be possible.

The image processor 300 is characterized in that filtering is performed using a Generalized-Cross-Validatory Spline Smoother (GCVSPS) algorithm.

Before analyzing measured data using motion capture, raw data should be filtered. The algorithm used for filtering uses the Generalized-Cross-Validatory Spline Smoother (GCVSPS) algorithm, which is commonly referred to as the Woltring filter. It is mainly used in the field of MA (Movement Analysis) and is known as the most accurate smoothing spline algorithm.

The difference from the Butter Worth Filter, which has been used for many years, can be applied even if the measurement sampling interval is different.

4 (a) and 4 (b) are graphs showing the results before and after the marker image filtering according to the present invention, respectively.

To compare the data before and after filtering, the stationary markers were measured for 2 minutes and the results were compared. 4 (a) shows the tendency of the movement of the marker with the low data before the filtering is applied, and FIG. 4 (b) shows the neat path compared to the data before the applying with the Waltling filter.

 Since the measuring markers are stationary, the coordinates should be stationary without movement. However, due to noise, lack of calibration, etc., the movements can be seen with a radius of 0.2 to 0.3 mm. It can be seen that.

The coordinate converter 400 includes a calibration unit 410 for setting an initial reference for the 3D coordinates and a coordinate calculator 420 for calculating 3D coordinates for the marker image based on the initial virtual 3D coordinates.

The calibration unit 410 sets the zero coordinates (0,0,0) as the reference for the three-dimensional coordinates, calculates the distance and direction of the camera and the zero coordinates, and sets the relationship between the cameras and the coordinates, The coordinate correction value may be determined in advance according to the type and performance.

Optical motion capture requires the process of forming virtual three-dimensional coordinates before measurement, which is called calibration.

Calibration should be performed for the following reasons. ① Set 3D spatial coordinates within the range that can be measured by motion capture equipment. At this time, the position of (0,0,0) can be set by the user. ② Calculate the distance and direction of cameras to measure and obtain the relationship between 3D space coordinates and each camera coordinate. ③ Correct the distorted coordinate information of the camera lens. Lenses used in motion capture equipment range from wide angle (8mm) to standard zoom, single lens and telephoto lens. In the case of the wide-angle lens, the distortion occurs as the distance from the center of the camera screen to the periphery, so it must be corrected through a software algorithm. Since there is no lens without distortion, it is necessary to apply an algorithm to identify and correct the distortion applied to each lens through calibration before measurement. ④ When the movement occurs in the installed camera, execute to regenerate the spatial coordinates. ⑤ Carry out accurate measurement by considering factors affecting motion capture equipment such as temperature, humidity, and ambient brightness of the place being measured.

Various methods are possible, but T-shaped wands of which the specifications are known are measured in advance, and coordinate calculation variables are modified so that the information obtained from the camera matches the T-shaped bar specifications. First, move the T-shaped bar well to all cameras for dynamic calibration. After measuring the movement, place the T-shaped bar where you want to generate the (0,0,0) coordinates and perform static calibration.

5 is a coordinate diagram for explaining a calibration process according to the present invention. The initial T-shaped bar is used to find the motion vector T and the rotation matrix R.

A simple pinhole camera model is used as shown in Figure 6 to obtain spatial three-dimensional coordinates from the two-dimensional coordinate information reflected on the camera screen. Figure 6 is largely divided into three coordinates-camera coordinates (C), CCD sensor coordinates (u), and spatial coordinates (W), and three-dimensionally using coordinates P ( x _u , y _u ) of the camera CCD sensor plane. The purpose is to find the space vector P _w .

O _C is the focus of the camera lens and the distance from the focus of the camera to the CCD sensor is f . Due to the structure of the camera, the CCD plane is located behind the focal point, but is shown to be located in front of FIG.

O _C , the focal point of the camera lens, is positioned at (o _x , o _y ) in the CCD plane along Z _C. Point P is an arbitrary point in three-dimensional space, expressed as a P _C vector based on camera coordinates, and expressed as a P _w vector based on spatial coordinates.

와

는 서로 닮음이고 z = f 이므로 아래의 수학식 1을 산출할 수 있다.

Wow

Since are similar to each other and z = f , the following Equation 1 can be calculated.

Where x _c , y _c , and z _c are the camera coordinates of the point where the P _C vector is bound to the camera CCD plane, and X _C , Y _C , and Z _C are the camera coordinates of the point P.

In order to convert the camera coordinate 'C' into the three-dimensional space coordinate 'W', a motion vector T and a rotation matrix R are required as shown in FIG. 5. The T vector represents the movement of O _C and O _W and the R matrix is a 3 × 3 orthogonal matrix. These two vectors are values obtained through the calibration process and can be used to express the relationship between the camera coordinates (C) and the three-dimensional coordinates (W) of the point P as follows.

i번째 행을 의미한다.
Where R _i is the R matrix

It means the i th row.

The relationship between a point P, the camera coordinates of the point P is projected on the CCD sensor and a CCD sensor coordinates of the three-dimensional space is equal to the expression (4) below.

Where h _x and h _y are the horizontal and vertical lengths of one pixel of the CCD sensor, respectively (pixel units are expressed in μm).

Using Equation 2 and Equation 4, a relationship between camera coordinates and three-dimensional coordinates for P may be calculated as follows.

Substituting Equation 6 into Equation 4 results in Equation 7 below.

Equation 8 below is used to expand two equations into three homogeneous equations.

In summary, Equation 8 may be expressed as Equation 9 below.

,

의 관계임을 알 수 있다. 따라서 (x _u , y _u )와 P _W 의 관계를 알 수 있다. 그러나 주의할 점은 선분 O _C P위의 모든 점은 CCD센서의 같은 한 점으로 상이 맺히게 된다. 따라서 상기 수학식 9에 스칼라 상수

를 곱하여 아래의 수학식 10과 같이 표현해 주어야 한다.
In Equation 8

,

It can be seen that the relationship. Therefore , we can know the relationship between (x _u , y _u ) and P _W. But notice that the line segment All the points on the O _C P are congruent with the same point on the CCD sensor. Therefore, the scalar constant in Equation 9

Multiply by to give it as shown in Equation 10 below.

값은 CCD로부터 P까지 거리를 알아야 구할 수 있다. 이러한 pin hole 카메라 모델은 3차원의 이미지가 2차원의 CCD 평면에 맺히는 관계식을 의미하며 켈리브레이션을 통해 얻어진 다수의 카메라 CCD의 정보를 바탕으로 3차원 좌표를 계산 하기 위해서는 추가적인 식이 필요하다. 모션 캡처에 사용되는 알고리즘에는 DLT(Direct Linear Transformation), NDLT(Non-Linear Direct Transformation) 등이 사용된다.a constant

The value can be obtained by knowing the distance from CCD to P. This pin hole camera model refers to the relational expression of three-dimensional image on the two-dimensional CCD plane, and additional equations are needed to calculate three-dimensional coordinates based on the information of a plurality of camera CCDs obtained through calibration. Algorithms used for motion capture include direct linear transformation (DLT) and non-linear direct transformation (NDLT).

The calibration unit 410 may be configured to reset an initial reference for three-dimensional coordinates when one or more cameras are changed in position. If the location of the camera has been changed due to a major shock or reinstallation in the camera's installed area, the calibration described above will need to be redone.

However, if only some of the cameras are repositioned, the repositioned camera may also be reset by using the relationship between the unaltered camera and the previous three-dimensional coordinates.

That is, if one or more cameras have not changed their positions when the positions of one or more cameras are changed, the calibrator 410 may determine whether the positions of the cameras and the coordinates of which the positions have been changed are based on the relationship between the cameras and the coordinates of which the positions have not changed. Calculate the relationship to reset the initial criteria for three-dimensional coordinates.

The coordinate calculator 420 calculates three-dimensional coordinates of the marker using a motion capture algorithm. In order to convert the camera coordinate 'C' into the three-dimensional space coordinate 'W', the above-described motion vector T and the rotation matrix R are required. That is, three-dimensional coordinates of the marker are measured by using the motion vector T and the rotation matrix R.

Hereinafter, a structure monitoring method using motion capture according to the present invention will be described. The overlapping description with the aforementioned structure monitoring apparatus will be omitted, and only the essential contents will be described.

6 is a flowchart illustrating a procedure of a structure monitoring method using motion capture according to the present invention.

Structure monitoring method using a motion capture according to the present invention is a step S1 to shoot a plurality of markers installed on the structure with a plurality of cameras, the initial reference of the three-dimensional coordinates for the structure is set, a plurality of cameras installed on the structure a plurality of markers In step S2 of filtering the image photographed by the S2, using the marker image filtered in the step S2 S3 of calculating the three-dimensional coordinates for the marker and S3 of the structure is measured by analyzing the three-dimensional coordinates calculated in the S3 step S4 Steps.

The marker is characterized in that it is installed on the structure according to the size, shape, predetermined deformation characteristics of the structure.

The marker is an active marker or a passive marker. In the case of the passive marker, an infrared or light generating device is disposed around the marker installed in the structure.

Step S2 according to the present invention is a structure monitoring method using motion capture, characterized in that the filter using the Generalized-Cross-Validatory Spline Smoother (GCVSPS) algorithm.

7 is a flowchart specifically showing step S1 of the present invention. In the step S1 according to the present invention, the step S1-1 in which zero coordinates (0,0,0), which is a reference to three-dimensional coordinates, is set in the image photographed by the camera, the camera is calculated by calculating a distance and a direction between the camera and the zero coordinates. And step S1-2 in which the relationship between the coordinates and the coordinates is set, and step S1-3 in which the coordinate correction value is determined according to the type and performance of the camera.

In the step S1 according to the present invention, when the position of one or more cameras among the cameras is changed, the initial reference for the 3D coordinates is reset.

In the step S1 according to the present invention, if there is at least one camera whose position has not changed when the position of one or more cameras among the cameras is changed, the position between the changed camera and the coordinates based on the relationship between the camera whose position has not changed and the coordinates. The initial criterion for the three-dimensional coordinates is reset by calculating the relation of.

Step S3 according to the present invention is characterized in that the three-dimensional coordinates for the marker are calculated using a motion capture algorithm.

Step S4 according to the present invention is characterized in that the three-dimensional strain, deformation acceleration and torsion of the structure are measured by analyzing the three-dimensional coordinates calculated in step S3.

The strain acceleration can be calculated through the numerical derivative of the displacement. Numerical differential formulas include forward, backward, and central finite difference formulas, but the central finite difference formula is the most accurate.

The embodiments and drawings attached to this specification are merely to clearly show some of the technical ideas included in the present invention, and those skilled in the art can easily infer within the scope of the technical ideas included in the specification and drawings of the present invention. Modifications that can be made and specific embodiments will be apparent that all fall within the scope of the present invention.

100: marker portion 200: camera portion
300: image processing unit 400: coordinate transformation unit
410: calibration unit 420: coordinate calculation unit
500: strain measurement unit

Claims (18)

A marker unit including a plurality of markers installed on the structure;
A camera unit including a plurality of cameras for photographing the markers;
An image processor filtering the marker images photographed by the plurality of cameras;
A coordinate converter configured to calculate three-dimensional coordinates of the marker using the marker image filtered by the image processor; And
Structure monitoring apparatus using a motion capture comprising a strain measuring unit for measuring the strain of the structure by analyzing the three-dimensional coordinates calculated by the coordinate transformation unit. The method of claim 1,
The marker of the marker unit is a structure monitoring apparatus using a motion capture, characterized in that the structure is installed on the structure according to the size, shape, predetermined deformation characteristics. The method of claim 1,
The marker may be an active marker or a passive marker. In the case of the passive marker, an infrared or light generating device is disposed around the marker installed in the structure. Monitoring device. The method of claim 1,
And the image processor performs filtering using a generalized-cross-validated spline smoother (GCVSPS) algorithm. The method of claim 1,
The coordinate transformation unit includes a calibration unit for setting an initial reference for the three-dimensional coordinates and a coordinate calculation unit for calculating the three-dimensional coordinates for the marker image based on the initial virtual three-dimensional coordinates Structure monitoring device. The method of claim 5,
The calibration unit sets zero coordinates (0,0,0) as reference for three-dimensional coordinates, calculates the distance and direction of the camera and zero coordinates, sets the relationship between the cameras and coordinates, and Structure monitoring apparatus using motion capture, characterized in that for determining the coordinate correction value according to the type and performance in advance. The method of claim 5,
The calibration unit structure monitoring apparatus using motion capture, characterized in that for resetting the initial reference to the three-dimensional coordinates, when one or more cameras are changed position. The method of claim 7, wherein
The calibration unit calculates the relationship between the coordinates of the changed camera and the coordinates based on the relationship between the coordinates of the camera and the coordinates that are not changed when one or more cameras are not changed when the positions of one or more cameras are changed. Structure monitoring apparatus using motion capture, characterized in that to reset the initial reference to the dimensional coordinates. The method of claim 5,
The coordinate calculation unit is a structure monitoring apparatus using a motion capture, characterized in that for calculating the three-dimensional coordinates for the marker using a motion capture algorithm. S1 step of photographing a plurality of markers installed on the structure with a plurality of cameras and the initial reference of the three-dimensional coordinates for the structure is set;
A step S2 of filtering images captured by the plurality of cameras of the plurality of markers installed in the structure;
S3 step of calculating the three-dimensional coordinates for the marker using the marker image filtered in the step S2; And
And a step S4 in which the strain of the structure is measured by analyzing the three-dimensional coordinates calculated in the step S3. The method of claim 11,
The marker is a structure monitoring method using a motion capture, characterized in that installed in the structure according to the size, shape, predetermined deformation characteristics of the structure. The method of claim 11,
The marker may be an active marker or a passive marker. In the case of the passive marker, an infrared or light generating device is disposed around the marker installed in the structure. Monitoring method. The method of claim 11,
The step S2 is a structure monitoring method using motion capture, characterized in that the filtering using the Generalized-Cross-Validatory Spline Smoother (GCVSPS) algorithm. The method of claim 11,
The step S1
Step S1-1 of setting zero coordinates (0,0,0) as a reference for three-dimensional coordinates in the image photographed by the camera;
Step S1-2 in which a relationship between the camera and coordinates is established by calculating a distance and a direction between the camera and zero coordinates; And
And a step S1-3 in which coordinate correction values are determined according to the type and performance of the camera. The method of claim 11,
The step S1
If the position of one or more of the cameras is changed, the structure monitoring method using a motion capture, characterized in that the initial reference for the three-dimensional coordinates is reset. 16. The method of claim 15,
The step S1
If there is one or more cameras whose position has not changed when the positions of one or more cameras are changed, the relationship between the coordinates of the changed camera and the coordinates is calculated based on the relationship between the cameras whose positions have not changed and the coordinates. A structure monitoring method using motion capture, characterized in that the initial reference to the dimensional coordinates are reset. The method of claim 11,
The step S3 is a structure monitoring method using a motion capture, characterized in that for calculating the three-dimensional coordinates for the marker using a motion capture algorithm. The method of claim 11,
In the step S4, by analyzing the three-dimensional coordinates calculated in the step S3, the three-dimensional strain, deformation acceleration and torsion of the structure is measured.

Images