JP3486239B2 - Orbital deviation measuring device and method, and curvature measuring method - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention continuously measures the deviation and height deviation and the curvature of a track after installation and track maintenance work under the condition that a predetermined track exists, such as a railway, using images. The present invention relates to an orbital deviation measuring device and method, and a curvature measuring method that can be performed.

[0002]

2. Description of the Related Art There are two types of rail deviations that are used as tracks: deviations and deviations in height. Passage refers to the amount of rail deviation in the direction perpendicular to the longitudinal direction of the rail on the track plane. Take an appropriate length, for example, 10 m, and connect a piano wire or the like between them to form a measurement string. When stretched, it is defined as a deviation of the distance between the measuring string and the rail (straight arrow) at the center point (straight arrow point) of the measuring string. Also,
Ups and downs are also defined to mean the ups and downs of the rails. When such a track deviation occurs, the vibration of the train increases, which makes the ride uncomfortable and makes high-speed traveling difficult. Therefore, track deviation is an important issue in the track of high-speed vehicles such as Shinkansen for high-speed running, and measurement of track deviation and its adjustment are very important tasks.

Conventionally, a track measuring vehicle has already been put into practical use as a device for measuring track deviation. There are various measuring methods by the measuring vehicle, but the deviation of the track is automatically measured by running on the track.

FIG. 16 illustrates a rail deviation and a conventional measuring method. FIG. 16A shows deviation when the rail 1 is vertical and deviation when the rail 1 is horizontal. Here, measurement points p and q forming an appropriate distance L of the rail 1
And the measurement string G is applied to both points, the measurement string G
The length Δs between the middle point and the corresponding point m of the rail 1 is defined as the deviation amount. However, in the vicinity of the curve part and the change point of the slope, the radius of curvature forms an arc, so the above Δs does not immediately show the amount of deviation, but the correct amount of deviation can be obtained by subtracting the displacement (straight arrow) due to the arc. To be

In order to obtain the above deviation amount Δs, it is necessary to properly set the reference line A and measure the displacement amounts of the three points p, q, and m with respect to it. FIG. 16 (b) shows a conventional measuring method in which the deviation in height is taken as an example, and the triaxial traveling wheel 3 mounted on the measuring vehicle 2 at an interval of ½ of the measurement string length L.
These wheels 3a, 3b, 3 are used by utilizing a, 3b, 3c.
The measurement points p, q, and m at three points on the rail 1 on both sides of c are made to correspond. Further, the height detection mechanisms 4a, 4b, 4c are attached to the axle boxes of the wheels 3a, 3b, 3c to set the reference line A of the measuring vehicle 2, and the traveling wheels 3a, 3b change with respect to the reference line A as the vehicle travels. , 3c, the heights Zp, Zq, and Zm can be measured at the same time, and the deviation amount ΔZ can be calculated by the following equation.

ΔZ = Zm− (Zp + Zq) / 2 (1) In the above, the height Z with respect to the three points p, q, and m
Although p, Zq, and Zm are measured at the same time, if these are measured at different points in time, the angle of the reference line A changes due to the rotation of the wheels during that time, so that the above-mentioned three running wheels 3a, Simultaneous measurement of 3 points by 3b and 3c is required. In addition, in the same way for passing, rail 1
3 sets of measuring wheels that come into contact with the vehicle and detect the left and right displacement thereof are provided, and the displacement amount Δy (y
Is perpendicular to the direction of the rail). However, when the measuring vehicle 2 is at a high speed, the measuring wheels interfere with running. Therefore, the Shinkansen measuring vehicle uses an optical displacement measuring device, and 3 sets of them are attached to the measuring vehicle to prevent wandering. Point measurement is done.

Further, there is a method using a laser beam as disclosed in Japanese Unexamined Patent Publication No. 5-213198, instead of a large-scale equipment such as the measuring vehicle 2. This method is composed of a light emitter and a light-receiving scale plate, and can measure in a longer span than the measuring wheel 2.

[0008]

However, it is difficult to connect the two devices, the light-emitting device and the light-receiving device together, and it is difficult to connect the devices, and a fixed distance is maintained for continuous measurement at the curved portion. , It is necessary to move them at the same time. Also,
In that case, at least two workers, a light emitting side and a light receiving side, are required. Further, in the conventional trajectory deviation measuring device, the deviation cannot be measured unless the straight arrow calculated by the curvature is subtracted from the measured amount as described above. Therefore, it is sufficient if the rails are laid out in a linear alignment, but if the laid rails do not have the correct curvature, there is a problem in that the amount of measurement does not immediately pass through at the curved portion.

The present invention has been made in view of the above-mentioned circumstances, and it is possible to easily measure a long span with a single measuring vehicle without installing two devices together. It is an object to provide an apparatus and method.
Another object of the present invention is to provide a curvature measuring method capable of automatically measuring the curvature at the curved portion of the trajectory.

[0010]

[Means for Solving the Problems]

(First Invention) A track deviation measuring device according to the present invention is a track for inspecting, when a rail to be a track is laid, whether or not the laid rail conforms to a plane alignment and a longitudinal alignment which are laying conditions. In the deviation measuring device, an image input means provided in a measuring vehicle traveling on the rail to two-dimensionally capture the rail, an image storage means for storing an image input by the image input means, and the measurement A height deviation measuring means mounted on a vehicle for measuring the height deviation of the rail, a position detecting means for detecting the current position of the measurement vehicle, and a data storage for preliminarily storing correct plane alignment and vertical alignment as data. Means, the amount of height deviation measured by the height deviation measuring means, the current position detected by the position detecting means, and the stored data of the data storage means Means for measuring the deviation of the rail of the actual image by drawing a rail image without deviation in the image obtained by the image input means, and a display device for displaying the deviation of the rail height and deviation of the rail measured by the above means And is provided.

(Second invention) A track deviation measuring method according to the present invention takes a rail which becomes a track at a place to be measured, stores the rail image in an image memory, and uses a displacement sensor attached to a measuring vehicle. The distance to the rail is measured continuously, the height deviation of the rail is calculated from the measured value, the above height deviation data is corrected by the previously entered laying data, and the correction vertical alignment is calculated. A rail in a normal state is drawn as a rail image by computer graphic technology from the input plane alignment and the corrected vertical alignment data, and the deviation amount of the rail is obtained by image processing from the rail image and the shooting rail image. To do.

(Third invention) The present invention relates to a curvature measuring method for measuring a curvature at a curved portion of a rail which is a track,
An image input means for providing a two-dimensional image of the rail, an image storage means for storing an image input by the image input means, which is provided in a measurement vehicle traveling on the rail, and mounted on the measurement vehicle. A height deviation measuring means for measuring the height deviation of the rail, a position detecting means for detecting the current position of the measuring vehicle, a data storage means for preliminarily storing correct plane alignment and vertical alignment as data, For the curved image of the rail taken by the image input means, sequentially draw the rail image with the curvature gradually changed, measure the overlapping state with the rail of the actual image, and measure the overlapping curvature with the actual rail curvature. It is characterized by making a judgment.

[0013]

When performing the measurement operation, first, the measurement start position is accurately input. Then, the rail of the place to be measured at that position is photographed by using an image input means provided on the measurement vehicle, for example, a two-dimensional image sensor, and the photographed rail image is stored in the image storage means. Next, the distance from the reference plane of the measurement vehicle to the rail is measured at three points by means of a height deviation measuring means, such as a displacement sensor, attached to the measurement vehicle, and the height deviation at this position is calculated from the measured data. This measurement of height variation is continuously performed while moving the measurement vehicle. At this time, the current position of the measurement vehicle is updated based on the rotation of the wheels and the like. The position data and the data of the vertical alignment (laying data relating to the height such as the slope) which is the installation data are corrected by the above measured height deviation amount, and the corrected vertical alignment is calculated and reconstructed.

Then, at the time when the height deviation of the image-capturing range can be measured first, that is, when the measurement target point is reached, computer graphic technology is applied,
Using the horizontal alignment that is the laying data and the corrected vertical alignment data that takes into account the height deviation that has been measured up to the local point, the rail in a normal state with no deviation as seen from the point where the rail image was first taken is used as the rail image above. Drawing in the image storage means. Then, this rail image is displayed so as to overlap the actual rail image, and the amount of shift at each point is measured by image processing. This shift amount is an incorrect amount immediately.

Then, the rail images photographed by the two-dimensional image sensor are continuously written in the image storage means and the above-mentioned measurement process is continuously executed, whereby the measurement at all points becomes possible.

Further, in the curved portion of the rail, the curvature itself may be abnormal. At this time, a rail image in which the curvature is changed little by little is created and compared with the actual rail image. The rail image curvature that matches well is obtained as the curvature at the curved portion.

By installing the track deviation measuring device on one measuring vehicle as described above, during track maintenance work, etc.
Can be easily operated by one person. In addition, if this measuring vehicle has a connecting function, it can be pulled by a track maintenance work vehicle or a confirmation vehicle for measurement, or it is possible to remove the complete device from the measurement vehicle and load it on the track maintenance work vehicle or confirmation vehicle. .

Further, upon measurement, the visual field range can be easily changed by changing the photographing magnification of the rail image, and can be freely set from a place with a large curvature to a place with a small curvature. Further, when measurement accuracy is required, the photographing magnification may be increased, and when measuring only the curvature, it is possible to select such as photographing a wide range with a low magnification. Furthermore, it is possible to perform measurements on curved sections that were impossible with conventional measurement vehicles.

[0019]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings. First, FIG. 1 shows a block diagram of a trajectory deviation measuring device of the present invention. This device is mounted on a measuring vehicle 10. A two-dimensional image sensor 11 that captures the front to measure the deviation of a rail serving as a track, an image memory 12 that stores the image, and this image. A display device for displaying the contents of the memory 12, for example, a monitor T with a touch panel
V13, a calculator 15 having data of track laying (planar alignment and longitudinal alignment) as a database 14, an interface 16 for inputting sensor signals, and first to third displacements for measuring rail height deviation The sensor 17a, 17b, 17c is provided. Further, a rotary encoder 19 for measuring a traveling distance is attached to a wheel 18 portion of the measuring vehicle 10. The rotary encoder 19 measures the traveling distance from the rotation of the wheels 18 and inputs it to the computer 15 via the interface 16. Then, by inputting the measurement start position and the traveling direction to the computer 15, the rotary encoder 1 is subsequently used.
The computer 15 can measure the current position of the measuring vehicle 10 based on the measurement signal of 9.

The plane linear laying data stored in the database 14 includes, for example, distinction between an up line and a down line, a range of distance from a reference position, a distinction between a straight line and a curve, a linear distance for a curve, a curve type, There are right and left curves, curve radius, curvature, etc. Further, the vertical line alignment data includes, for example, distinction between an up line and a down line, a range of a kilometer from a reference position, a type of line, a grade of a gradient, a gradient radius, an altitude, and the like. Then, the installation data stored in the database 14 is read according to the current position of the measuring vehicle 10 obtained by the computer 15 from the measurement data of the rotary encoder 19.

FIG. 2 shows a schematic configuration diagram when the track deviation measuring device shown in FIG. 1 is used by being mounted on a measuring vehicle 10. In FIG. 2, an example is shown in which the measuring vehicle 10 is towed by a track maintenance work vehicle and moved to the site, and the measuring vehicle 10 is separated and used alone at a measurement location.
In this case, the worker may push the measurement vehicle 10 to perform the measurement, or the engine may be attached to the measurement vehicle 10 and self-propelled to perform the measurement. The measurement vehicle 10 is provided with wheels 18a and 18b so that it can travel on a track, and a rotary encoder 19 is attached to one of the wheels 18a. Further, on the lower side of the measuring vehicle 10, for example, optical first to third displacement sensors 17a to 17c for measuring the height deviation of the rail 1 are arranged at the center and front and rear positions along the reference line A at the center and front and rear positions. It is installed with the interval of / 2. The detection signals of these displacement sensors 17a to 17c are all input to the computer 15 via the interface 16 (see FIG. 1), and the height deviation of the rail 1 is required.

For measurement of deviation, a two-dimensional image sensor 11 for taking a front rail image, an image memory 12 for storing the image, various measurement information, data stored in the database 14 and the like are used. There is provided a calculator 15 for measuring the deviation of the vehicle, and a monitor TV 13 with a touch panel for displaying and inputting the operation information and the deviation of the trajectory. The computer 15 is a rotary encoder 1
Database 14 based on the position information measured by 9
In addition to having a function of reading plane linear or longitudinal linear data stored in advance and drawing a rail image on the image memory 12 using the computer graphic technique without any track deviation, the data is also stored in the image memory 12. It has a function to measure the deviation of the rail image by extracting only the rail of the front rail image and measuring the deviation between the rail of the actual image and the rail image.

Next, the procedure for measuring the trajectory deviation will be described with reference to the flowchart of FIG. The measurement vehicle 10 is moved to the measurement start position, data indicating the measurement start position, for example, the distance from the reference position is input to the calculator 15, and then the traveling direction, that is, the up direction or the down direction is input ( Step A1). In this case, a technique such as measuring the distance from the point detector or other laying objects or the position using a satellite may be used. Then, the two-dimensional image sensor 11 captures an image of the rail 1 in front of it, that is, at a place to be measured, and the captured rail image is stored in the image memory 12 (step A2). Then, the movement of the measuring vehicle 10 is started to measure the deviation in height (steps A3 and A4).

Here, the displacement sensors 17a to 17a arranged along the lower reference line A of the measuring vehicle 10 are used to calculate the height deviation.
First, the sensor 1 in the center of the measuring vehicle
The distance Zm to the rail 1 is continuously measured by 7a. Similarly, the displacement sensors 17b and 17c continuously measure the distances Zp and Zq to the rail 1 before and after the measuring vehicle, and the height deviation is calculated using the equation (1). However, when this device is mounted on a large vehicle such as a work vehicle and used, the distance L / 2 between the displacement sensors 17a, 17b, and 17c can be set sufficiently, but when carrying in the track maintenance work, the measurement vehicle 10 The sensor distance L
/ 2 cannot be taken sufficiently. So in this case,
As shown in FIG. 3 (a), instead of the displacement sensors 17b and 17c, two slit light sources 20a and 2 are provided in front of and behind the measuring vehicle 10.
0b and two-dimensional cameras 21a and 21b are provided, and the measurement vehicle 1
A device such as measuring the displacement of the rail 1 at a point away from 0 is required. In this case, the displacement sensor 17a at the center of the measuring vehicle is used as it is. The slit light source 2
0a, 20b irradiates the rail 1 in front of a few meters with the light 22 in the form of a thin slit of laser light, and the two-dimensional cameras 21a, 21b mounted on the rail 1 at a position higher than the slit light sources 20a, 20b. Take a picture of the slit light. An example of the image at this time is shown in FIG. In this case, the light sources 20a and 20b are slit-shaped because the image of the rail 1 is not a straight line as shown in FIG. Is. At this time, the two-dimensional cameras 21a, 21
If the intersection of the rail 1 and the slit light 22 on the image photographed at b comes to the center line 23 of FIG. 3B, it is the same as the distance Zm obtained by the center displacement sensor 17a, that is, the horizontal plane B. Thus, the two-dimensional cameras 21a and 21b and the light sources 20a and 20b are set. In this way, the height differences Zp and Zq can be measured by measuring the amounts Dp and Dq at which the intersections of the rail 1 and the slit light 22 deviate from the center line 23 on the screens of the two-dimensional cameras 21a and 21b.

As this method, the height differences Zp and Zq are previously set.
And the center line 23 on the image, rail 1 and slit light 22
The height difference Zp, Zq can be easily measured by checking the relationship between the displacement points Dp, Dq with the intersection point of. In this way
The distances Zm, Zp, and Zq at three points can be obtained, and the height deviation can be measured from the above formula (1). This has been described with respect to a portion which is not a vertical curve, but on the vertical curve, the deviation amount can be obtained by taking into account the displacement amount previously calculated from the vertical alignment.

Returning to FIG. 6, the explanation of the measuring means will be continued. Although the rail height deviation is measured as described above, it is judged whether or not the measurement is completed (step A
5) If the measurement is not completed, the measurement operation of step A4 is continued. Then, when the measurement target point is reached and the above measurement operation is completed, the measurement results of the rail height deviation are totaled (step A6). Next, the vertical deviation data that is the laying data is corrected using the deviation amount obtained as described above, and the vertical alignment of the current rail, that is, the corrected vertical alignment is reconstructed by calculation (step A7). ). Therefore, since the position information from the position where the image was first captured, the plane alignment, and the corrected vertical alignment are known, the line shape without the height deviation in the front seen from that position can be known. In other words, the track shape (rail image) of the front rail, which is visible from a certain position and has no deviation in height, is drawn in the image memory 12 and the monitor TV 1 is displayed.
3 can be displayed on the screen.

Specifically, as shown in FIG. 4A, three-dimensional coordinates O having the origin at the track center portion at the head of the measurement vehicle and the rail head surface.
With the -XYZ system, the coordinates p (x, y, x) of an arbitrary rail can be calculated using the installation data (planar alignment and vertical alignment). When the front of the two-dimensional image sensor 11 of the measurement vehicle is photographed, the position (m, n) where the point p is visible on the screen as shown in FIG. 4B is given by mathematical coordinate conversion. In order to explain this coordinate conversion, a coordinate system as shown in FIG. 5 is set. (1) Coordinates whose origin is the rail central surface at the center of the track at the beginning of the O-XYZ system measurement vehicle. (2) The two-dimensional image sensor 11 is installed at the origin O'of the O'-X'Y'Z 'system, and its optical axis is the Y'axis. This coordinate system is a two-dimensional image sensor 1
The coordinate system is determined by the mounting condition of No. 1, and the O-XYZ system is translated in parallel by the mounting height h and the deviation w from the center, and rotation components α (yaw), β (pitch), and γ (roll) around each axis. ) Is added. (3) The image plane of the o'-x'y'z 'system two-dimensional image sensor 11 is represented by o'-x'y'z'.
The origin of the O'-X'Y'Z 'system is O'
It is translated by f along the Y ′ axis. (4) The o-MN system o'-x'y'z 'is a discretized coordinate system and is set so that o'is the center of the screen. The coordinates actually viewed on the screen are data in this coordinate system. Then, the coordinates (m, n) are obtained by performing the following coordinate conversion. (1) Conversion from O-XYZ system to O'-X'Y'Z 'system A combination of parallel movement and rotation gives the following equation.

[0028]

[Equation 1] (2) Conversion from the O'-X'Y'Z 'system to the o'-x'y'z' system This is a similar conversion proportional to the parallel movement amount f and is represented by the following equation.

[0029]

[Equation 2] (3) Conversion from o'-x'y'z 'system to o-MN system The horizontal angle of view of the two-dimensional image sensor 11 is 2θh, the number of pixels is M, and the vertical angle of view is 2θv. Number of pixels is N
And If the image resolutions are gm and gn, respectively, the following equation holds.

Gm = 2ftan θh / M gn = 2ftan θv / N (4) Add an offset for the origin o ′ to come to the center of the screen,

[0031]

[Equation 3]

By the above, the coordinates (m, n) on the screen are obtained. m = M / 2 (1 + X '/ Y'tan θh) n = N / 2 (1 + X' / Y'tan θv) (6) Therefore, the rail image ahead by an arbitrary distance is stored in the image memory 12. It can be drawn and displayed on the screen of the monitor TV 13 (step A8). FIG. 8 shows an example of the rail image 30a in the straight line portion.

The rail image 30a thus obtained
Is used as a "measurement" to measure orbital deviation. However, there may be a deviation between the rail image 30a and the actual rail image extracted from the captured image of the two-dimensional image sensor 11. FIG. 9 shows an example of the actual rail image 31a extracted from the captured image of the two-dimensional image sensor 11. In addition, FIG.
Shows the image shift when the rail image 30a and the actual rail image 31a are superimposed.

The deviation between the rail image 30a and the actual rail image 31a as described above is because the measuring vehicle 10 is not parallel to the track or meanders. Also,
This is because it was displaced in the vertical and horizontal directions due to vibration. Therefore,
The rail image 30a is moved in parallel in the vertical and horizontal directions to perform template matching with the actual rail image 31a so that the rail image 30a and the actual rail image 31a overlap each other to correct the deviation,
Passing is measured (steps A9 and A10).

Next, a method for template matching (hereinafter referred to as rail matching) between the actual rail image 31a and the rail image 30a and a method for extracting the rail image 31a from the imaged image of the two-dimensional image sensor 11 for measurement of deviations will be described. To do. (1) Utilizing the feature that the rails are continuous in the vertical direction of the screen, an image (moving average image) fL (x, of the background part in which the moving average of the original image f (x, y) is calculated in the horizontal direction is calculated.
y) is calculated.

[0036]

[Equation 4] (2) The moving average image fL (x, from the original image f (x, y)
y) is subtracted to obtain a difference image fH (x, y).

FH (x, y) = f (x, y) −fL (x, y) (8) (3) The density of the partial image of the difference image fH (x, y) is changed. An average value (μ) and a standard deviation (σ) are calculated and binarized by the following formula to obtain a binarized image g (x, y).

G (x, y) = 1; fH (x, y) ≧ μ + Kσ 0; fH (x, y) <μ + Kσ (9) However, K: constant A binary value obtained in this way Template matching is performed between the image g (x, y), that is, the actual rail image 31a and the rail image 30a, the position having the strongest correlation is obtained, and the positional relationship of the measuring device and the trajectory are corrected.

FIG. 11 shows a state in which the rail image 30a is drawn at a position having the strongest correlation with the actual rail image 31a. Then, as shown in FIG. 12, the shift amount T in the left-right direction from the actual rail image 31a is measured with each point of the rail image 30a as a reference. The value of T becomes a deviation amount at each point (step A11).

The above-mentioned method is an embodiment in which only the image is considered, but the two-dimensional image sensor 11 measures the sensor and the vertical vibration so that the two-dimensional image sensor 11 is in a predetermined position and parallel to the trajectory. If a sensor is attached and the posture is controlled to be correct (parallel to the rail), it is possible to perform measurement without performing matching.

Then, the rail images taken by the two-dimensional image sensor 11 are continuously written in the image memory 12,
By continuously executing the above-described measurement processing, it is possible to measure the trajectory deviation at all points.

Next, a description will be given of a case where the curvature is obtained when the curvature itself is different and the curvature itself is different. When obtaining the curvature of the rail 1, the image memory 1
Although it is possible to obtain by the inverse operation of the above calculation from the rail on the image stored in No. 2, here, the case of obtaining using plane linear data will be described. Figure 7
It is a flow chart which shows the procedure.

Basically, it becomes the same as the trajectory deviation, but the rail image is drawn by gradually increasing or decreasing the curvature value R obtained from the plane linear data, and the rail matching with the actual rail image is performed. Curvature R'that maximizes the matching rate
Ask for. Further, as the type of curve, for example, a circular curve,
Although there is a relaxation curve, etc., a rail image is drawn based on planar linear data.

First, the curvature measurement start position is input to the computer 15, and then the traveling direction, that is, the up direction or the down direction is input (step B1). Next, the front rail image is taken by the two-dimensional image sensor 11 and the image memory 1
It is stored in 2 (step B2), the movement is started, and the height deviation is measured (steps B3, B4). Then, it is judged whether or not the measurement of the rail height deviation is completed (step B5), and if the measurement is not completed, the measurement operation of step B4 is continued. Then, after the measurement operation is completed,
The measurement results of rail height deviation are totaled (step B6).
Then, using the height deviation obtained above, correct the vertical alignment that is the laying data to obtain the vertical alignment of the current rail,
That is, the corrected vertical alignment is reconstructed by calculation (step B
7). The processing of steps B1 to B7 is performed in the same manner as the processing of steps A1 to A7 in FIG. Then, first, in the curved portion of the rail 1, an actual rail image is extracted from the captured image of the two-dimensional image sensor 11, and the extracted actual rail image 31b is passed through the image memory 12 as shown in FIG. Drawing on the screen of the monitor TV 13.

Next, the type of the curve, the curvature R, etc. are read from the plane linear data stored in the database 14 (step B8), and the changing curvature ΔR for gradually changing the curvature is initially set to "0". , The curvature R'of the rail image to be drawn is obtained by "R '= R + ΔR" (steps B9, B10). Then, using this curvature R ', a rail image 30b of the type shown by plane linear data is drawn on the screen as shown in FIG. 14 (step B1.
1) Overlay with the actual rail image 31b (step B12). Next, rail matching is performed between the rail image 30b and the actual rail image 31b (step B13), and what percentage of the entire drawn rail image 30b is matched (coincidence rate) is calculated (step B14). Next, it is checked whether or not this matching rate is the maximum, and if it is not the maximum, "1" is added to the changing curvature ΔR in step B16, the process returns to step B10, and the above-mentioned processing is repeated.

As described above, the rail image 30b is made by gradually changing the curvature R ', and the rail matching with the actual rail image 31b is performed to find out what percentage of the entire drawn rail image is matched. Repeat. Then, as shown in FIG. 15, the most matching curvature R'is set as the actual curvature of the rail (step B17). This method has an advantage that it can be realized by an extension line of rail image drawing without complicated calculation. The method of obtaining the shift amounts Dp and Dq of the rail 1 described in the above embodiment is not limited to the method of the above embodiment.
Various modifications can be made without departing from the spirit of the present invention.

[0047]

As described above, according to the present invention, even a single worker can efficiently measure the track deviation for a continuous rail by a simple operation. Also,
The measurement accuracy can be achieved by changing the image capturing magnification. Further, from a case where the curvature is small to a case where the curvature is large, the trajectory deviation can be measured relatively easily within the range that can be overlooked. Further, according to the curvature measuring method of the present invention, the curvature itself at the curved portion of the trajectory can be measured reliably and easily.

[Brief description of drawings]

FIG. 1 is a block diagram of a trajectory deviation measuring device according to an embodiment of the present invention.

FIG. 2 is a schematic configuration diagram showing a state in which the track deviation measuring device according to the embodiment is mounted on a measuring vehicle.

FIG. 3A is a configuration diagram in the case of performing the height deviation measurement when the measurement vehicle size is small, and FIG. 3B is a diagram showing an example of an image when the height deviation measurement is performed in the above (a).

4A is a schematic perspective view for explaining a rail image calculation method, and FIG. 4B is a schematic view showing positions on a two-dimensional image sensor corresponding to FIG. 4A for explaining a rail image calculation method. Explanatory drawing.

FIG. 5 is a diagram showing coordinate systems for calculating a rail image.

FIG. 6 is a flowchart showing a method of measuring orbital deviation.

FIG. 7 is a flowchart showing a curvature measuring method.

FIG. 8 is a diagram showing an example of a rail image in a straight line portion.

FIG. 9 is a diagram showing an example of an actual rail image extracted from a captured image of a two-dimensional image sensor in a straight line portion.

FIG. 10 is a diagram showing an image example in which a rail image and an actual rail image are superimposed.

FIG. 11 is a diagram showing an example of an image as a result of performing template matching.

FIG. 12 is a diagram for explaining a method of measuring a deviation amount T.
The figure which expands and shows a part of 1.

FIG. 13 is a diagram showing an example of an actual rail image extracted from a captured image of a two-dimensional image sensor in a curved portion.

FIG. 14 is a diagram showing a rail image of a curvature R at a curved portion.

FIG. 15 is a diagram showing a rail image of a curvature R + ΔR at a curved portion.

FIG. 16A is an explanatory diagram of a rail deviation and a conventional measurement method, and FIG. 16B is an explanatory diagram of a conventional measurement method in which height deviation is an example.

[Explanation of Codes] 10 Measuring Vehicle 11 Two-dimensional Image Sensor 12 Image Memory 13 Monitor TV 14 Database 15 Computer 16 Interfaces 17a to 17c Displacement Sensors 18, 18a, 18b Wheels 19 Rotary Echoders 20a, 20b Slit Light Sources 21a, 21b 2 Dimensional camera 22 Slit light emitted from the front rail 23 Center lines 30a, 30b of the image Rail images 31a, 31b Actual rail image extracted from the image captured by the two-dimensional image sensor

Front page continued (72) Inventor Norihiro Ichikura 2-6 Yoyogi, Shibuya-ku, Tokyo Inside Safety Research Institute, East Japan Railway Company (72) Inventor Atsushi Sasaki 2-6 Yoyogi, Shibuya-ku, Tokyo East Japan Railway Company Safety Research Institute (72) Hiroshi Yamashita Hiroshi Yamashita 4-22 Kannon Shinmachi, Nishi-ku, Hiroshima City, Hiroshima Prefecture Mitsubishi Heavy Industries, Ltd. Hiroshima Research Institute (72) Atsushi Nakamoto Kannon, Nishi-ku, Hiroshima City, Hiroshima Prefecture 4-6-22 Shinmachi, Mitsubishi Heavy Industries, Ltd. Hiroshima Research Laboratory (56) Reference JP-A-6-249658 (JP, A) JP-A-5-116626 (JP, A) JP-A-5-157548 (JP, A) JP-A-6-42948 (JP, A) JP-A-5-151341 (JP, A) JP-A-6-94442 (JP, A) (58) Fields investigated (Int.Cl. ⁷ , DB name) ) G01C 7/00 B61K 9/08 G01B 11/00 G01B 21/00 G01C 15/00

Claims (3)

(57) [Claims] 1. A track deviation measuring device for inspecting whether or not a laid rail conforms to a laying condition such as a plane alignment and a vertical alignment when the laid rail is laid, and runs on the rail. An image input means provided on the measuring vehicle for two-dimensionally photographing the rail, an image storing means for storing an image input by the image input means, and an image mounting means mounted on the measuring vehicle for adjusting the height of the rail. High and low deviation measuring means, position detecting means for detecting the current position of the measuring vehicle, data storage means for storing the correct plane alignment and vertical alignment in advance as data, and the high and low deviation measuring means for measuring Based on the amount of deviation of height, the current position detected by the position detecting means, and the stored data of the data storing means, within the image obtained by the image inputting means. A track having a means for measuring the deviation of the rail of the actual image by drawing a rail image without deviation and a display device for displaying the deviation of the rail height and deviation of the rail measured by the above means. Crazy measuring device. 2. A rail which becomes a track at a place to be measured is photographed, the rail image is stored in an image memory, a distance to the rail is continuously measured by a displacement sensor attached to a measuring vehicle, and the measured value is measured. Calculate the rail height deviation from the above, correct the above height deviation data by the previously entered laying data, and calculate the corrected vertical alignment. From the previously input plane alignment and the corrected vertical alignment data at the measurement target point, normal state The rail deviation measurement method is characterized in that the rail is drawn as a rail image by computer graphic technology, and the rail deviation is calculated by image processing from the rail image and the above-mentioned shooting rail image. 3. A curvature measuring method for measuring a curvature at a curved portion of a rail serving as a track, comprising image input means provided in a measuring vehicle traveling on the rail, for two-dimensionally photographing the rail, An image storage means for storing an image input by the image input means, a height deviation measuring means mounted on the measuring vehicle for measuring height deviation of the rail, and a position detecting means for detecting a current position of the measuring vehicle. Data storage means for storing the correct plane alignment and vertical alignment in advance as data, and the rail image with the curvature gradually changed with respect to the curved portion image of the rail photographed by the image input device described above. A curvature measurement method characterized by measuring the overlapping state of a real image with a rail and determining the most overlapping curvature as the actual rail curvature.