JP6061205B2 - Method and apparatus for constantly monitoring structural displacement - Google Patents

Description

  The present invention has a plurality of targets at a displacement monitoring point of a structure such as a road bridge, and these structures are digitized by photographing these targets with a camera or the like, and the structure relative displacement is performed using the data. The present invention relates to a continuous monitoring method, a GPS sensor installed in a structure, and a constant monitoring method and apparatus for the displacement of the entire structure using a reception signal of a field side device.

Various monitoring methods and devices of this type have been known (Patent Documents 1, 2, and 3).
In Patent Document 1, the trajectory is tilted from an oblique direction by an electronic imaging means installed with the optical axis oriented obliquely with respect to the extending direction of the trajectory of the measurement target section so that a plurality of signs are all included in one image. The technique of photographing is described. It also describes that the sign is illuminated at night, and that the sign is entirely black and has a white pattern in the center.

  Japanese Patent Application Laid-Open No. 2004-228561 describes a technique that enables accurate measurement by obtaining and subtracting an error due to movement even when the camera device moves.

  Patent Document 3 discloses a displacement measurement device that calculates a displacement amount at an arbitrary measurement point from images before and after displacement of a plurality of measurement objects, and the image processing unit is configured to select an arbitrary one from a plurality of measurement points of an image. Select a measurement point as a reference point, calculate the displacement of each measurement point based on the reference point, select the reference point from the previously captured coordinate value and the change amount of the coordinate value captured later, and perform any measurement A technique is described in which, when it is determined that the amount of change in the coordinate value of a point is smaller than a predetermined value, such a measurement point is used as a reference point.

JP 2004-053374 A JP 2007-240218 A JP 2008-216158 A

The invention described in Patent Document 1 requires a calibration work as a preparatory work in order to accurately measure a trajectory error, and measures a displacement point that must be performed by a complicated three-dimensional coordinate system calculation. The calculation method for this is extremely complicated.
The invention described in Japanese Patent Application Laid-Open No. H10-228561 requires a calculation with an extremely complicated mathematical expression to obtain an error due to movement of the camera device.
In the invention described in Patent Document 3, the three-dimensional coordinates of the measurement point before displacement and the three-dimensional coordinates of the measurement point after displacement measured when the first photographed image and the second photographed image are superimposed. The amount of displacement of the measurement point is calculated by obtaining the difference between the first image and the three-dimensional coordinates. In order to obtain the three-dimensional coordinates, the first image taken from the left side, the second image taken from the right side, the shooting position and the shooting direction of the camera This method also has a problem in that this method requires extremely complicated equipment and computation.

  The present invention uses a camera such as an existing road management camera, can measure the relative displacement of a structure with a simple calculation, avoids the influence of noise at the time of target detection, and can measure with high accuracy. It is an object of the present invention to provide a method and apparatus for constantly monitoring structure displacement.

The present invention measures displacement by a simple vector method, and even if the camera rotates or rotates during monitoring and the camera orientation and zoom magnification change, the displacement can be easily corrected and measured. This is one of the features.
In the present invention, the displacement monitoring location has a plurality of targets. The target may be one that can be identified by the unique shape and color of a structure that already exists without self-luminous, such as a road marking line, or one that is self-luminous and always lights without flashing. Although it is good, a target that flashes and emits light itself is desirable in order to be resistant to noise caused by external light and to be able to shoot at night. When using a blinking target, at least three images that are turned on and off in the same time width and with a period slightly shorter than half of the blinking period are taken, and the target is detected by differential processing of these images. The method.
In the case of using a self-luminous target, it is also effective to employ an infrared light emitting target for noise reduction and to attach an infrared transmission filter to the camera.
If the target to be monitored rotates due to the displacement of the structure, the effect of template matching will deteriorate. In the present invention, a circular target is adopted so that the displacement point can be detected stably even when the target rotates. ing.
If the displacement measurement target is farther than the distance between at least two fixed reference targets, the error in the estimated coordinates of the displacement measurement target increases in proportion to the distance, and relative displacement measurement with high accuracy is achieved. However, in the present invention, the sub-pixel technique of the template matching method is adopted to measure the relative displacement with high accuracy.
According to the present invention, by combining a GPS sensor with an image processing method and apparatus, it is possible to measure displacement such as movement of the entire bridge that cannot be clearly understood only by measuring some relative displacement such as a bridge girder.

According to invention of Claim 1,
At least two reference targets for reference are provided at locations where displacement is unlikely to occur, and at least one displacement measurement target is provided at the displacement monitoring location, and these targets are photographed with a camera to digitize the image. In the method of constantly monitoring the displacement of the structure to measure and constantly monitor the displacement of the displacement point using the data,
Initializing seeking each measurement start coordinates of the at least two particular said the two reference target of one of the at least one specific one of the displacement measurement target of the image of the camera Process,
The image coordinates are translated so that the coordinates of one of the two reference targets at the time of measurement by the camera of the reference target become the coordinates of the measurement start of the one reference target. The measurement start of the other reference target of the two specific reference targets with respect to the coordinate of the one reference target after the parallel movement, the relative coordinates at the time of measurement, and the one specific displacement measurement The process of obtaining the relative coordinates at the start of measurement of the target for measurement ,
A step of determining the estimated coordinates of the displacement measurement target on the basis of said relative coordinates of the measurement start the relative coordinates and the displacement measuring target during measurement and the measurement start of the target for the other reference,
A step of calculating a displacement amount of the displacement measurement target based on the estimated coordinates of the displacement measurement target and the relative coordinates at the time of measurement of the displacement measurement target. Since the monitoring method is adopted, the relative displacement of the structure can be constantly monitored by calculating the relative displacement amount of the displacement measurement target using a simple vector method. Further, when the displacement measurement target due to the displacement of the structure rotates, the rotation angle of the target can be measured when a non-circular target is employed. However, when a non-circular target is adopted, the target coordinate detection efficiency by template matching is deteriorated. However, in the present invention, a circular target is adopted, and the target can be efficiently detected without being affected by the rotation of the target. .

According to invention of Claim 4,
At least two reference targets at locations where displacement is unlikely to occur;
At least one displacement measurement target at the displacement monitoring point;
A camera for photographing these reference targets and displacement measurement targets and constantly monitoring them,
A computer as an image processing unit for image signals sent from the camera,
This computer,
Initializing seeking each measurement start coordinates of the at least two particular said the two reference target of one of the at least one specific one of the displacement measurement target of the image of the camera means,
The image coordinates are translated so that the coordinates of one of the two reference targets at the time of measurement by the camera of the reference target become the coordinates of the measurement start of the one reference target. The measurement start of the other reference target of the two specific reference targets with respect to the coordinate of the one reference target after the parallel movement, the relative coordinates at the time of measurement, and the one specific displacement measurement Means for obtaining the relative coordinates at the start of measurement of the target for measurement ,
Means for obtaining an estimated coordinates of the displacement measurement target on the basis of said relative coordinates of the measurement start of the relative coordinates and the displacement measuring target during measurement and the measurement start of the target for the other reference,
A computer-readable recording recording a program for functioning as a means for calculating the amount of displacement of the displacement measurement target based on the estimated coordinates of the displacement measurement target and the relative coordinates at the time of measurement of the displacement measurement target Since the medium is provided, existing devices such as a road management camera and an optical fiber communication network can be used to the maximum, and can be used at a low cost and in many places.

According to invention of Claim 5,
The target is turned on and off in the same time width, and the sampling cycle t2 is within the range of 0.5 × t1 <t2 <1.0 × t1, where t1 is the blinking cycle of the target. By acquiring three images in succession, the difference images of the first and second sets, the second and third sets are obtained in these images, and these differences are obtained. Target coordinates are detected by using at least one set of lit and unlit images in the image, so there are many noises that affect shooting, such as car lights in outdoor environments. Even so, the coordinates of the target can be reliably detected by processing the two difference images simultaneously.

According to the sixth aspect of the present invention, when a target that emits infrared rays is used as a target and a filter that transmits only infrared rays is attached to the camera, various visible light noises can be cut, and the coordinates of the target are obtained. Can be reliably detected.

According to the eighth aspect of the present invention, since the target is formed in a circular shape, the presence of the natural world is small and it is easy to identify the target, and the influence of the installation requirements can be minimized.

It is a side view of the bridge girder which shows one Example of the continuous monitoring method of the structure displacement by this invention, and its apparatus. It is a top view in FIG. An example of the displacement of a bridge girder is shown. (A) is a plan view of a state where the bridge girder is laterally displaced, (b) is a plan view of a state where the bridge girder is displaced vertically, and (c) is a case where the bridge girder is displaced vertically. A side view of the state, (d) is a side view of a state in which the road surface on the back of the abutment has been undulated such as depression. (A) is explanatory drawing of the coordinate of the target in the original position (position before rotation etc.) of a camera, (b) is the coordinate of the target when a reference point is displaced with zooming, rotation, etc. of a camera. It is explanatory drawing of. It is explanatory drawing which detects the displacement amount of a displacement point from the coordinate of a target. It is a top view of the target used by this invention. (A) is explanatory drawing of the 1st example which forms a difference image from the input image at the time of using a blink target, (b) is explanatory drawing of the 2nd example which forms a difference image from an input image. It is explanatory drawing which shows the relationship between a polar coordinate and a rectangular coordinate. It is explanatory drawing explaining a subpixel method. It is a block diagram which shows one Example of the constant monitoring method and the apparatus of the structure displacement by this invention. It is a flowchart which shows one Example of the continuous monitoring method of the structure displacement by this invention, and its apparatus.

The present invention has at least two reference targets 15a at locations where displacement is unlikely to occur, and has at least one displacement measurement target 15b at locations where displacement is monitored. It is a method of constantly monitoring the displacement of a structure to digitize the image and use the data to measure the displacement at the displacement point and constantly monitor it.
Detecting the coordinates of the reference target 15a and the displacement measurement target 15b at the beginning of measurement and at the time of measurement from the image of the camera 14;
Obtaining estimated coordinates of the displacement measurement target 15b from the coordinates of the reference target 15a at the beginning of measurement and the coordinates of the displacement measurement target 15b at the beginning of measurement;
Calculating the amount of displacement of the displacement measurement target 15b from the estimated coordinates of the displacement measurement target 15b at the time of measurement and the coordinates of the displacement measurement target 15b detected by the image of the camera 14 at the time of measurement.

The target 15 forms a donut-shaped self-light-emitting portion 19 that emits light at the center of the light-shielding plate 21 that reflects little light and cannot transmit light on the back surface, and the central portion of the self-light-emitting portion 19 is a circular light absorbing portion. It is characterized by being 20.
The target 15 blinks, and when the blinking cycle of the target 15 is t1, the sampling interval t2 of the one-time image capturing is 0.5 × t1 <t2 <1.0 × t1. By capturing at least three images in succession within the range, a difference image of each of the first and second sets, the second and third sets is obtained in these images, The coordinate detection of the target 15 is performed using the fact that the difference image includes the lit image and the unlit image of each target.

  The target may be a target that can be identified by a unique shape or color of the structure without self-light emission, such as a road marking line, or a target that is self-light-emitting and always lights without blinking.

  1 and 2 are a side view and a plan view showing displacement monitoring of a structure such as a road bridge, for example, a bridge girder 12, and an abutment 11 whose upper end is flush with the road surface 10 on both banks of a river or the like. It is assumed that the bridge girder 12 is mounted via the support 13. Between the both ends of the bridge girder 12 and the abutment 11, a clearance 16 for adjusting the amount of expansion / contraction of the bridge girder 12 is usually provided.

A camera 14 is installed to constantly monitor whether or not the bridge girder 12 which is such a structure is displaced due to an earthquake, flood, landslide or the like.
The present invention uses such an existing road management camera 14 and corrects the structure displacement correctly by performing correction processing on the image even if the angle of view or zoom magnification of the camera 14 is displaced during monitoring. It is characterized by monitoring. This camera 14 may be a fixed dedicated camera for the purpose of the present invention of constantly monitoring structure displacement.

At least two reference targets 15a serving as reference points are provided on the side of the road surface 10 that is not displaced or hardly generates displacement. It is desirable that the reference target 15a is firmly attached so that the interval does not change. Further, at least one displacement measurement target 15b is provided on the bridge girder 12 which is a measurement object.
The present invention measures the displacement of the displacement measuring target 15b relative to the reference target 15a on the assumption that the mutual distance and altitude of the reference target 15a are not displaced, and the reference target 15a and the displacement measuring target are measured. Although the target 15b is arranged on a straight line in the illustrated example, the target 15b is not necessarily arranged on a straight line.
Further, in order to measure the undulation of the road surface 10, the reference target 15 a may be provided at a stable point such as the abutment 11, and the displacement measurement target 15 b may be provided on the road surface 10 that is easily displaced.

  Using at least two reference targets 15a as described above as reference points, the displacement of at least one displacement measurement target 15b is image-processed and measured. The displacement includes a lateral displacement 17a as shown in FIG. 3A, a longitudinal displacement 17b as shown in FIG. 3B, a step 17c as shown in FIG. 3C, and a undulation 17d as shown in FIG. There are complex displacements, and these displacements are measured by the method and apparatus described later.

As shown in FIG. 1 and FIG. 2, at the location to be measured, at least two reference targets 15 a and at least one displacement measurement target 15 b are provided and a camera 14 is installed. 3 (a), (b), (c), and (d) in a plurality of targets 15 from one captured surface image by photographing a plurality of targets 15 from an oblique direction with a single camera. It can be measured as a two-dimensional displacement shown in FIG.
As shown in FIG. 10, the image data taken by the camera 14 is transmitted and received to the main processing unit 23 such as a management office using an existing optical fiber communication network 25 for road management or the like. As a result, maintenance costs can be significantly reduced.
In FIG. 10, an image signal can be transmitted from the transmission unit 24 on the camera 14 side constituting the site side apparatus 22 to the management office side constituting the main processing apparatus 23.
The received data is sent to the image processing unit 27 of the camera having a computer function. The image processing unit 27 of the camera performs input target signal filtering processing, blinking signal processing, template matching processing, color / brightness processing, combination processing thereof, and the like. These processed signals are sent from the output unit 28 to the memory 29, the display device 30, and the alarm device 31.

  A GPS sensor 33 is installed in or near the on-site device 22 composed of the camera or the like, and this GPS signal is sent to the GPS data processing unit 32 via the receiving unit 26 of the main processing unit 23 such as a management office. By connecting to the display device 30, the memory 29, and the alarm device 31 via the feed and output unit 28, the displacement of the entire structure that cannot be discriminated by the image or calculation of the target 15, such as the ground of the bridge moving, etc. It is also possible to measure and issue an alarm.

Based on FIG. 11, the operation of the constant displacement monitoring method and apparatus according to the present invention will be described.
(1) Preparation work for image processing is performed at the initial installation of the system.
The following processing conditions a) to c) are set for automatic measurement by the structure displacement constant monitoring method and apparatus of the present invention.
a) Registration of template Using the sample image, the template image of each target is registered.
b) Registration of search area The search range of each target is registered using the sample image.
c) Adjustment of image processing parameters ・ Setting of parameters for calculating target displacement, etc.

(2) Detection of target coordinates by image processing The method and apparatus for constantly monitoring structure displacement according to the present invention are repeatedly executed in the following procedure.
a1) When using a target that is not self-luminous, such as a road marking line, can be identified by the unique shape or color of the structure, or is self-luminous and always lights without flashing, etc. Follow the steps.
Read one image.
The search area image is cut out for the captured image.
a2) On the other hand, when a blinking target is used, the following procedure is executed.
-Read at least three consecutive images.
The difference processing of the first and second groups of the three images and the second and third groups are performed to represent the two difference images.
-The search area image is cut out for two difference images.
Hereinafter, the procedure that does not blink and the procedure that blinks are the same.
(3) Target coordinate detection is performed on the image in the search area.
Since it is essential to reliably detect the coordinates of the target in an outdoor environment and is specific to the present invention, details will be described later.
(4) Repeat (2) and (3) until there is no next search area.
(5) Determine the coordinate detection status of the target.
(6) When correctly detected, the amount of displacement on the target image is calculated.
Using the target coordinates of the initial image at the start as the target coordinates at the beginning of the measurement and the coordinates of the target detected from the subsequent images as the coordinates of the target at the time of measurement, the displacement amount of the target for displacement measurement is obtained. This step is unique to the present invention, and details will be described later.
(7) It is determined whether the displacement amount exceeds a warning value.
(8) If the warning value is exceeded, a warning message is output.
(9) Return to (2) and repeat the process.

  The road management camera 14 has a main purpose of road management, and therefore, in response to a command from the management office, the camera or the like is rotated to take a surrounding image or zoom to enlarge the image. In addition, the road management camera 14 has a preset function (a function for storing a shooting location, a range, and the like in advance), but the accuracy when returning to the same situation with the preset function is not necessarily high. The coordinates of the target 15 on the image vary depending on the zoom and rotation angle of the camera 14. Further, the rotation angle of the camera 14 may change due to road vibration or the like.

Considering the above points, as shown in FIGS. 1 and 2, the targets 15 a and 15 b are arranged so that one or a plurality of sets of the reference target 15 a and the displacement measurement 15 b fit within the image of one camera 14. Is set.
Next, the correction of the reference points of the targets 15a and 15b and the measurement of the displacement amount will be described.
4A is an image before correction, and FIG. 4B is an image in which an error due to zooming, rotation, or the like of the camera 14 occurs, at least two reference images on these images are used. By using the condition that the physical position of the target 15a is constant, it is possible to correct an error caused by zooming or rotation of the camera 14.

The calculation of the displacement amount from the image coordinates of the target 15 will be described.
A plurality of targets 15 are used for measuring the displacement of the target 15 in order to cope with a change in the shooting angle of the camera 14 and a change in zoom.
In FIG. 5, the physical position between T ₁₀ and T ₂₀ of the reference target 15a is constant, and the relative displacement amount of the target 15 is measured by measuring the relative displacement amount of T ₃₀ of the displacement measurement target 15b. To do.

The image processing unit 27 of the camera 14 is a computer itself, and includes a computer-readable recording medium that records a program for causing the computer to function as each unit described below.
Specifically, the present invention causes a computer as the image processing unit 27 of the camera 14,
Means for initially setting coordinates on an image corresponding to reference coordinates at the start of measurement of each of the reference targets 15a and the displacement measurement targets 15b;
Means for obtaining relative polar coordinates between the start of measurement of the reference target 15a and the time of measurement from the image of the camera 14;
Means for obtaining estimated polar coordinates of the displacement measurement target 15b from relative polar coordinates at the time of measurement of the reference target 15a;
Means for obtaining relative polar coordinates of the displacement measurement target 15b from the image of the camera 14 at the time of measurement;
The displacement measurement target 15b is determined from the estimated polar coordinates of the displacement measurement target 15b obtained from the relative polar coordinates of the reference target 15a at the time of measurement and the relative polar coordinates of the displacement measurement target 15b obtained from the image of the camera 14 at the time of measurement. Means for calculating the displacement amount of
A computer-readable recording medium that records a program that functions as a computer.

The procedure for calculating the displacement amount from the polar coordinates of the target 15 by the above means will be described below.
(1) The initial coordinates (T ₁₀ (x ₁₀ , y ₁₀ ), T ₂₀ (x ₂₀ , y ₂₀ ), T ₃₀ (x ₃₀ , y ₃₀ )) of each target 15a, 15b are registered as initial setting values. To do.
(2) It is assumed that an error has occurred between the beginning of measurement and the state at the time of measurement due to a change in the rotation of the camera 14, zooming, etc. before the processing target image is captured.
(3) The coordinates on the image of T ₁₁ at the time of measurement are (x ₁₁ , y ₁₁ ), and the image coordinates are translated so that (x ₁₁ , y ₁₁ ) becomes (x ₁₀ , y ₁₀ ). The relative coordinates on the image of T ₁₁ , T ₂₁ , T ₃₁ at the time of measurement after the parallel movement of the image coordinates are (x ₁₀ , y ₁₀ ), (x ₂₁ , y ₂₁ ), (x ₃₁ , y ₃₁ ). .
(4) T ₁₀ and T ₂₀ are connected by a rigid body, the distance R is constant, and the relative polar coordinates at the beginning of measurement with respect to T ₁₁ of T ₂₁ and the relative polar coordinates at the time of measurement are (R ₂₀ , ρ ₂₀ ) Coordinates) and (R ₂₁ , ρ ₂₁ ) (measurement coordinates).
(5) The relative polar coordinates at the beginning of measurement and the relative polar coordinates at the time of measurement of T ₃₀ with respect to T ₁₀ are (R ₃₀ , ρ ₃₀ ) (coordinates at the beginning of measurement) and (R ₃₁ , ρ ₃₁ ) (coordinates at the time of measurement).
(6) T ₃₂ calculates the displacement amount on the image using the following formula, assuming that the estimated polar coordinates on the image at the time of measurement are (R ₃₂ , ρ ₃₂ ).

Under the above conditions,
R ₃₂ = R ₃₀ × (R ₂₁ / R ₂₀ )
ρ ₃₂ = ρ ₃₀ + ρ ₂₁ −ρ ₂₀
For this reason, the estimated displacement amounts Δx ₃ and Δy ₃ on the screen of T ₃₀ can be expressed by the following equations.
Δx ₃ = R ₃₁ × cos ρ ₃₁ −R ₃₂ × cos ρ ₃₂ Equation 1
Δy ₃ = R ₃₁ × sin ρ ₃₁ −R ₃₂ × sin ρ ₃₂ Formula 2
become. here,
R ₂₀ = √ {(x ₂₀ −x ₁₀ ) ² + (y ₂₀ −y ₁₀ ) ² }
ρ ₂₀ = tan ⁻¹ {(y ₂₀ −y ₁₀ ) / (x ₂₀ −x ₁₀ )}
R ₃₀ = √ {(x ₃₀ −x ₁₀ ) ² + (y ₃₀ −y ₁₀ ) ² }
ρ ₃₀ = tan ⁻¹ {(y ₃₀ −y ₁₀ ) / (x ₃₀ −x ₁₀ )}
R ₂₁ = √ {(x ₂₁ −x ₁₁ ) ² + (y ₂₁ −y ₁₁ ) ² }
ρ ₂₁ = tan ⁻¹ {(y ₂₁ −y ₁₁ ) / (x ₂₁ −x ₁₁ )}
R ₃₁ = √ {(x ₃₁ −x ₁₁ ) ² + (y ₃₁ −y ₁₁ ) ² }
ρ ₃₁ = tan ⁻¹ {(y ₃₁ −y ₁₁ ) / (x ₃₁ −x ₁₁ )}
It is.

(7) The measured movement amounts ΔX ₃ , ΔY _3, and ΔZ ₃ of T ₃₀ can be expressed by the following equations.
ΔX ₃ = Δx ₃ × A _x + Δy ₃ × B _x Formula 3
ΔY ₃ = Δx ₃ × A _y + Δy ₃ × B _y Formula 4
ΔZ ₃ = Δx ₃ × A _z + Δy ₃ × B _z Formula 5
become. Here, A _X , B _X , A _y , B _y , A _z , and B _z are actually measured values of the displacement / pixel in each direction of one pixel on the T ₃₀ image at the initial setting.
(8) The displacement amount / pixel of one pixel at the initial setting is a value adjusted according to the actual environment based on the value shown in FIG.

In the coordinates shown in FIG. 5, the calculation example using the polar coordinates is shown, but the present invention is not limited to this, and the coordinates may be based on the orthogonal coordinates.
FIG. 8 shows an example of conversion between polar coordinates and rectangular coordinates. Since the coordinate point (x, y) is represented by polar coordinates (R, ρ) and orthogonal coordinates are represented by (R cos ρ, R sin ρ). Mutual coordinates can be easily converted.

Adoption of the flashing target 15 In an outdoor environment, there are many noises that affect shooting, such as car lighting. Further, at night, when the target 15 does not emit light, it cannot be recognized from the captured image.
In order to solve this problem, in the present invention, the coordinate detection of the target 15 is performed using the difference data of the blinking image of the target 15.
In actual operation, no synchronization processing is performed between the blinking cycle of the target 15 and the sampling cycle of the camera 14. Furthermore, there is an error in both cycles. For this reason, depending on the sampling timing of the images, the continuously captured images may be turned off at the same time or turned on at the same time, and the target 15 cannot be detected correctly in the image difference processing.
In order to solve this problem, the sampling of the camera 14 captures at least three images continuously at a period different from the blinking period of the target 15. Specifically, it is assumed that the time widths of turning on and off of the target 15 are the same, and the sampling period t2 of shooting by the camera is set to a range of 0.5 × t1 <t2 <1.0 × t1 of the blinking period t1 of the target. Capture at least three consecutive images. Then, a differential image of each of the first and second sets, the second and third sets is obtained from these images, and at least one set of the lighting image and the non-lighting image is included in these difference images. The coordinates of the target are detected by using

FIG. 7A shows the first and second images in FIG. 7A when the first and second images of at least three consecutive input images are lit images and the third image is an unlit image. The difference between the sets of eyes is an unlit image, and the second and third sets are turned on images.
In FIG. 7B, when the first image of at least three consecutive input images is a lit image and the second and third images are unlit images, the first image in FIG. The difference between the second set and the second set is a lit image, and the second and third set is a non-lit image.
Thus, by processing two difference images simultaneously, the coordinates of the target 15 can be reliably detected.

The target 15 is determined in consideration of the following points.
(1) A shape that allows easy image processing.
・ By adopting a few shapes in the natural world, it becomes easy to specify the target.
In the case of target detection by the pattern matching method, a circular shape is desirable in order to minimize the influence of installation requirements such as target rotation.
(2) A shape with high target detection accuracy.
-Since coordinates are detected in units of subpixels, it is desirable that the target to be detected has a clear pattern and edges in each direction.
Considering the above points, as shown in FIG. 6, a donut-shaped self-light-emitting portion 19 that emits red light, such as an LED, is formed in the center of a square non-reflective light-shielding plate 21. Is a circular light absorber 20. The inner periphery and the outer periphery of the self-light-emitting portion 19 are provided with a black border.
The target 15 is not limited to the above example, and may be one that emits light other than red, one that does not emit light but is coated with a highly reflective material, one that constantly emits light instead of blinking, and the like.

The two images captured as the difference images can be detected stably even when the target 15 rotates by detecting the target 15 by a known pattern matching method.
In addition, the template matching method of sub-pixel technology can be used to measure displacement with higher accuracy. This will be described with reference to FIG.
The image of the camera 14 is in units of pixels, but when obtaining the coordinates of the object, the detection accuracy is one pixel.
On the other hand, in the normalized correlation search, a value of one pixel or less of coordinates can also be output by calculation using a nearby correlation value. Although the case where a quadratic curve is used is illustrated below, other functions may be used.
In FIG. 9, when the coordinate in the X direction of the maximum point (P _x ) of the detected correlation value is x and the correlation values of the left and right coordinates are P _x−1 and P _{x + 1} ,
P _x-1 = A (x-1-B) ² + C
P _x = A (x−B) ² + C
P _{x + 1} = A (x + 1−B) ² + C
From the above formula, the coordinate in the X direction at the peak of the quadratic curve is B, and B is expressed by the following formula.
_{X max = B = x + (} P x + 1 -P x-1) / 2 (2P x -P x-1 -P x + 1)

The plurality of targets installed at the displacement monitoring point are preferably a target that flashes and emits light in order to be resistant to noise due to external light and to be able to shoot at night.
However, the target may be one that can be identified by the unique shape and color of the existing structure without self-luminous, such as a road marking line, or it is self-luminous and always lit without flashing It may be a thing to do.

In the embodiment, since the reference target 15a is two and the displacement measurement target 15b is one, the relative displacement amount of the displacement measurement target 15b is measured.
However, the present invention is not limited to this, and by setting two displacement measurement targets 15b at a predetermined interval, the distance between the two displacement measurement targets 15b is based on the distance between the two reference targets 15a. By calculating the relative displacement, the relative displacement amount can be verified, and more reliable measurement can be performed.

  DESCRIPTION OF SYMBOLS 10 ... Road surface, 11 ... Abutment, 12 ... Bridge girder, 13 ... Support, 14 ... Camera, 15 ... Target, 16 ... Spacing, 17 ... Displacement point, 19 ... Self-luminous part, 20 ... Light absorption part, 21 ... Light-shielding plate , 22 ... field side device, 23 ... main processing device, 24 ... transmitting unit, 26 ... receiving unit, 27 ... image processing unit of camera, 28 ... output unit, 29 ... memory, 30 ... display device, 31 ... alarm device, 32: GPS data processing unit, 33: GPS sensor.

Claims (8)

At least two reference targets for reference are provided at locations where displacement is unlikely to occur, and at least one displacement measurement target is provided at the displacement monitoring location, and these targets are photographed with a camera to digitize the image. In the method of constantly monitoring the displacement of the structure to measure and constantly monitor the displacement of the displacement point using the data,
Initializing seeking each measurement start coordinates of the at least two particular said the two reference target of one of the at least one specific one of the displacement measurement target of the image of the camera Process,
The image coordinates are translated so that the coordinates of one of the two reference targets at the time of measurement by the camera of the reference target become the coordinates of the measurement start of the one reference target. The measurement start of the other reference target of the two specific reference targets with respect to the coordinate of the one reference target after the parallel movement, the relative coordinates at the time of measurement, and the one specific displacement measurement The process of obtaining the relative coordinates at the start of measurement of the target for measurement ,
A step of determining the estimated coordinates of the displacement measurement target on the basis of said relative coordinates of the measurement start the relative coordinates and the displacement measuring target during measurement and the measurement start of the target for the other reference,
A step of calculating a displacement amount of the displacement measurement target based on the estimated coordinates of the displacement measurement target and the relative coordinates at the time of measurement of the displacement measurement target. Monitoring method. The target is turned on and off in the same time width, and the sampling cycle t2 is within the range of 0.5 × t1 <t2 <1.0 × t1, where t1 is the blinking cycle of the target. By acquiring three images in succession, the difference images of the first and second sets, the second and third sets are obtained in these images, and these differences are obtained. 2. The method for constantly monitoring the displacement of a structure according to claim 1, wherein the coordinates of the target are detected by utilizing that the image includes at least one set of lighting image and light extinguishing image. At least two reference targets for reference are provided at locations where displacement is unlikely to occur, and at least one displacement measurement target is provided at the displacement monitoring location, and these targets are photographed with a camera to digitize the image. In the method of constantly monitoring the displacement of the structure to measure and constantly monitor the displacement of the displacement point using the data,
The polar coordinates (R ₁₀ , ρ ₁₀ ) and (R ₂₀ , ρ ₂₀ ) at the start of measurement of the specific two of the at least two reference targets from the image of the camera and the specific of the at least one Obtaining and initial-setting polar coordinates (R ₃₀ , ρ ₃₀ ) at the beginning of measurement of one displacement measurement target;
Of the two specific reference targets, polar coordinates (R ₁₀ , ρ ₁₀ ) at the time of measurement of one reference target with the camera are used as polar coordinates (R ₁₁ , ρ ₁₁ ) is performed to translate the image coordinates, and the other reference of the two specific reference targets with respect to the polar coordinates (R ₁₁ , ρ ₁₁ ) of the one reference target after the translation is performed. Relative polar coordinates (R ₂₀ , ρ ₂₀ ) at the start of measurement of a target for measurement, relative polar coordinates (R ₂₁ , ρ ₂₁ ) at the time of measurement, and relative polar coordinates (R ₃₀ , ρ ₂₁ ) at the start of measurement of the one specific displacement measurement target ρ ₃₀ ) and relative polar coordinates (R ₃₁ , ρ ₃₁ ) at the time of measurement ;
The relative polar coordinates (R ₂₀ , ρ ₂₀ ) at the start of measurement of the other reference target, the relative polar coordinates (R ₂₁ , ρ ₂₁ ) at the time of measurement, and the relative polar coordinates (R _{30 at the} start of measurement of the displacement measurement target ). , Ρ ₃₀ ) to obtain the estimated polar coordinates (R ₃₂ , ρ ₃₂ ) of the displacement measurement target from the following equation:
R ₃₂ = R ₃₀ × (R ₂₁ / R ₂₀ )
ρ ₃₂ = ρ ₃₀ + ρ ₂₁ −ρ ₂₀
Based on the estimated polar coordinates (R ₃₂ , ρ ₃₂ ) of the displacement measurement target and the relative polar coordinates (R ₃₁ , ρ ₃₁ ) at the time of measurement of the displacement measurement target, the displacement amount (Δx _3, Δy of the displacement measurement target ) ₃ ) calculating from the following equation:
Δx ₃ = R ₃₁ × cosρ ₃₁ −R ₃₂ × cosρ ₃₂
Δy ₃ = R ₃₁ × sinρ ₃₁ −R ₃₂ × sinρ ₃₂
A method for constantly monitoring a displacement of a structure, comprising: At least two reference targets at locations where displacement is unlikely to occur;
At least one displacement measurement target at the displacement monitoring point;
A camera for photographing these reference targets and displacement measurement targets and constantly monitoring them,
A computer as an image processing unit for image signals sent from the camera,
This computer,
Initializing seeking each measurement start coordinates of the at least two particular said the two reference target of one of the at least one specific one of the displacement measurement target of the image of the camera means,
The image coordinates are translated so that the coordinates of one of the two reference targets at the time of measurement by the camera of the reference target become the coordinates of the measurement start of the one reference target. The measurement start of the other reference target of the two specific reference targets with respect to the coordinate of the one reference target after the parallel movement, the relative coordinates at the time of measurement, and the one specific displacement measurement Means for obtaining the relative coordinates at the start of measurement of the target for measurement ,
Means for obtaining an estimated coordinates of the displacement measurement target on the basis of said relative coordinates of the measurement start of the relative coordinates and the displacement measuring target during measurement and the measurement start of the target for the other reference,
A computer-readable recording recording a program for functioning as a means for calculating the amount of displacement of the displacement measurement target based on the estimated coordinates of the displacement measurement target and the relative coordinates at the time of measurement of the displacement measurement target A structure displacement monitoring device characterized by comprising a medium. The target is turned on and off in the same time width, and the sampling cycle t2 is within the range of 0.5 × t1 <t2 <1.0 × t1, where t1 is the blinking cycle of the target. By acquiring three images in succession, the difference images of the first and second sets, the second and third sets are obtained in these images, and these differences are obtained. 5. The apparatus for constantly monitoring a displacement of a structure according to claim 4, wherein the coordinates of the target are detected by using that the image includes at least one set of lighting image and non-lighting image. An output device coupled to the output side of the image processing unit, combines the GPS sensor output installed in the structure, according to claim 4 or 5 further characterized in that the displacement of the entire structure was measurable structure A constant monitoring device for object displacement. 6. The apparatus for constantly monitoring displacement of a structure according to claim 5 , wherein the target emits infrared light, and an infrared transmission filter is attached to the camera. 6. The structure displacement monitoring device according to claim 5 , wherein the target is circular.

Images