KR101414370B1 - Train car position measurement device for electric railway maintenance - Google Patents

Abstract

An apparatus for measuring the position of an electric railway repairing vehicle, which is capable of specifying a relationship between a traveling position of a vehicle and a passing time with high accuracy, comprising: a line sensor (1) An image creating section 10 for creating an image taken by the sensor 1 as a template image and an input image, a memory 11 for storing a template image and an input image created by the image creating section 10, A template creation processing unit 13 for extracting a SIFT feature amount from the template image and creating template data and storing the SIFT feature amount in the memory 12 and a SIFT feature amount extraction unit 13 for extracting the SIFT feature amount from the input image and storing the SIFT feature amount in the memory 12 A processing unit (14) for detecting a position of the hanger in the input image by searching for a correspondence point between the SIFT characteristic amount of the stored template data and the SIFT characteristic amount of the input image, Re was found to having a hanger position detection processing unit 15 for retention (12).

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention [0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an electric railway repair vehicle position measuring apparatus for measuring a traveling position of an electric railway repairing vehicle used for repairing an overhead line and a track, And acquires the relationship between the traveling position and the passage time.

Electric railway facilities are mainly wire and track. These are important repair facilities for each railway. While the electric railway vehicle is operated, the electric wire contacts the electric power collecting device every time the electric railway vehicle passes through it. For this reason, the wire is gradually worn, and if it is not exchanged, the wire is finally broken and may cause an accident. In addition, warpage, warping, scratches, abrasion, and the like are caused by the operation of the tram, and if the progress of the trajectory is left unintentionally, the trajectory may cause an accident such as derailment of the electric railway vehicle.

In this regard, in order to quickly identify and respond to a critical point caused by abrasion of the wire or trajectory, it is necessary to determine the position of the critical point of the wire or track (for example, Etc.) is becoming an important issue in terms of maintenance.

In the case of repairing a facility such as a wire or a track, a repair-only vehicle or a track-and-land vehicle is currently used. As a result of the detection, if a point of interest is detected, the position of the point of interest (See, for example, Patent Document 1 below).

Here, as a method of specifying the position of the vehicle, for example, there is a method using an automatic train control device (ATC) or a rotary encoder, and each has the following features.

The automatic train control device controls the position and speed of a train and is originally a system for driving a vehicle. However, this automatic train control device has a mechanism for communicating information on the position and speed of the train to the outside. Therefore, when specifying the vehicle position using the automatic train control device, the position of the repair-only vehicle is specified by using a mechanism that communicates with the outside of the automatic train control device (for example, refer to Patent Document 2 below) .

When the vehicle position is specified using the rotary encoder, it is possible to determine, based on the number of revolutions of the wheel obtained by the rotary encoder, the number of revolutions of the wheel and the circumferential length of the wheel, (For example, refer to Patent Document 3 below).

Japanese Laid-Open Patent Publication No. 2008-89523 Japanese Patent Application Laid-Open No. 2006-248412 Japanese Laid-Open Patent Publication No. 2004-17700

 Fujiyoshi Hironobu, "Gradient-Based Feature Extraction-SIFT and HOG-", Information Processing Society Research Report CVIM 160, 2007, p.211-224

However, in the case of using the above-described conventional automatic train control device, in addition to the provision of the transmission / reception devices on the vehicle side, it is necessary to install the transmission / reception devices on the trajectory side at regular intervals. There is a problem in that the number is increased and the cost is increased.

Since the rotary encoder is configured to measure the travel distance by counting the rotational angle of the wheel, the train is not traveling due to braking or running on the slope, and the wheel is rotating, or due to slippage of the wheel There is a problem that locally generated errors such as the progress of a train are increasing additionally while the wheel is not rotating, thereby making it difficult to specify the position of the critical point.

Therefore, in a conventional automatic train control device, a line sensor camera (hereinafter referred to as a line sensor) is provided on the roof of the vehicle, and an image near the line taken by the line sensor is subjected to image processing Detecting the hanger, detecting the hanger, and specifying the relationship between the traveling position and the passing time of the vehicle based on the time at which the hanger was photographed.

However, if the hanger itself hangs obliquely or if the hanger is inclined on the screen, if the hanger is detected by template matching, the accuracy of hanger detection during collision with the template may deteriorate. In addition, There is a possibility that the accuracy of hanger detection is deteriorated in that the longitudinal edge is weakened.

From the above, it is an object of the present invention to provide an electric railway repair vehicle position measuring apparatus capable of accurately specifying a relationship between a traveling position and a passing time of a vehicle.

According to a first aspect of the present invention for solving the above problems, there is provided an apparatus for measuring the position of a vehicle for repairing an electric railway,

A line sensor for capturing an image of a hanging line for hanging a hanger supporting a wire or a wire,

An image creating unit that creates an image captured by the line sensor as a template image and an input image;

A storage unit that stores the template image and the input image created by the image creation unit;

A template creation processing unit for extracting SIFT feature amounts from the template image stored in the storage unit and creating template data and storing the template data in the storage unit,

An SIFT feature extraction processor for extracting a SIFT feature quantity from the input image stored in the storage and storing the SIFT feature quantity in the storage,

And a hanger position detection processing unit for searching a corresponding point between the SIFT feature amount of the template data stored in the storage unit and the SIFT feature amount of the input image, detecting the hanger position in the input image, and storing the detected hanger position in the storage unit .

In order to solve the above-mentioned problems, an electric railway repair vehicle position measuring apparatus according to a second invention is the electric railway repair vehicle position measuring apparatus according to the first invention,

And a time recording processing unit for calculating the time at which the hanger is detected based on the sampling frequency of the line sensor and the position of the hanger on the input image stored in the storage unit and storing the time data in the storage unit .

In order to solve the above-mentioned problem, an electric railway repair vehicle position measuring apparatus according to a third invention is the electric railway repair vehicle position measuring apparatus according to the first invention or the second invention,

Wherein when the previous hanger position is present in the hanger position detection processing unit, the previous hanger position and the inclination data of the hanger and the crossover are received and the next processing zone And a calculation is performed.

In order to solve the above-mentioned problems, an electric railway repair vehicle position measuring apparatus according to a fourth invention is the electric railway repair vehicle position measuring apparatus according to the third invention,

And in the hanger position detection processing section, the next processing area is extended.

According to the present invention, it is possible to provide an electric railway repair vehicle position measurement apparatus capable of specifying the relationship between the traveling position of the vehicle and the passage time with high accuracy.

1 is a schematic view showing a configuration of an electric railway repair vehicle position measuring apparatus according to a first embodiment of the present invention.
2 is a block diagram showing the configuration of an electric railway repair vehicle position measuring apparatus according to the first embodiment of the present invention.
3 is a flowchart showing a processing procedure in the vehicle position-measuring apparatus for electric railway repair according to the first embodiment of the present invention.
4 is a diagram showing an example of a template image in an electric railway repair vehicle position measuring apparatus according to the first embodiment of the present invention.
5 is a diagram showing an example of the SIFT characteristic quantity of the template data in the electric railway repair vehicle position measuring apparatus according to the first embodiment of the present invention.
6 is a diagram showing an example of an input image in an electric railway repair vehicle position measuring apparatus according to the first embodiment of the present invention.
7 is a diagram showing an example of a correspondence point search processing region of the SIFT characteristic amount in the electric railway repair vehicle position measuring apparatus according to the first embodiment of the present invention.
8 is a block diagram showing a configuration of an electric railway repair vehicle position measuring apparatus according to a second embodiment of the present invention.
Fig. 9 is a flowchart showing a processing procedure in an electric railway repair vehicle position measuring apparatus according to the second embodiment of the present invention.
Fig. 10 is a diagram showing an example of a next processing area in an electric railway repair vehicle position measuring apparatus according to the second embodiment of the present invention.
11 is a view showing an example in which a next processing area in an electric railway repair vehicle position measuring apparatus according to the third embodiment of the present invention is expanded.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, an embodiment of an electric railway repair vehicle position measuring apparatus according to the present invention will be described with reference to the drawings.

Example 1

Hereinafter, a first embodiment of an electric railway repair vehicle position measuring apparatus according to the present invention will be described.

The electric railway maintenance repair vehicle position measuring apparatus according to the present embodiment uses a line sensor provided on the roof of a vehicle to photograph a line and a line with the line sensor facing the sleeping direction and with the slope upward, A SIFT (Scale-Invariant Feature Transform) feature quantity (see the non-patent reference 1) is obtained as an image feature quantity which is not influenced by a change in rotation, and a correspondence point with a SIFT feature quantity extracted from a previously- And the hanger is detected.

First, an overview of an electric railway repair vehicle position measuring apparatus according to the present embodiment will be described.

1 is a schematic diagram showing an overview of an electric railway repair vehicle position measuring apparatus according to the present embodiment.

1, an electric railway repair vehicle position measuring apparatus according to the present embodiment is provided with a hanger 102 for picking up an image of a hanger 103 holding a hanger 101 supporting a hanger 101 or a hanger 101 A line sensor 1 which is installed with a treading direction on the roof of the vehicle 100 so as to be inclined upward and which takes an image on a photographing line indicated by an arrow L in Fig. 1, A light source 2 installed in the vicinity of the line sensor 1 and a vehicle 100 for storing an image photographed by the line sensor 1 and performing image processing And a processing PC 3 that is installed in the processing system 1. The line sensor 1 and the processing PC 3 are connected so that image data can be transmitted. The direction indicated by the arrow A in Fig. 1 indicates the traveling direction of the vehicle 100. [

Next, the construction of an electric railway repair vehicle position measuring apparatus according to the present embodiment will be described.

2 is a block diagram showing a configuration of an electric railway repair vehicle position measuring apparatus according to the present embodiment.

2, an electric railway repair vehicle position measuring apparatus according to the present embodiment includes an image creating section 10 for creating an image photographed by the line sensor 1 as a template image and an input image, A memory 11 for storing a template image and an input image created by the creating unit 10, a template 11 for extracting a SIFT feature amount from the template image stored in the memory 11, creating template data, A SIFT feature amount extraction processing unit 14 for extracting a SIFT feature amount from the input image stored in the memory 11 and storing the extracted SIFT feature amount in the memory 12, A hanger position detection processing section (15) for searching a corresponding point between the SIFT characteristic amount and the SIFT characteristic amount of the input image to detect the position of the hanger in the image and storing it in the memory (12) And a wave number and a memory recording the time processing unit (16) for retention on the basis of the position on the image saved in the hanger 12 and the time data to calculate the a hanger 102, detection time in the memory 12.

In this embodiment, the processing PC 3 is connected to the image creating unit 10, the memories 11 and 12, the template creation processing unit 13, the SIFT feature extraction processing unit 14, the hanger position detection processing unit 15, a hanger position detection processing section 15 and a time recording processing section 16. [ The memory 11 and the memory 12 may be a common memory.

Next, the procedure of the electric railway repair vehicle position measuring apparatus according to the present embodiment will be described.

Fig. 3 is a flowchart showing a processing procedure in an electric railway repair vehicle position measuring apparatus according to the present embodiment.

As shown in Fig. 3, as a processing procedure in the vehicle position determining system for an electric railway repair according to the present embodiment, firstly, a template image is acquired from the line sensor 1 in Step P10. Further, as shown in Fig. 4, the template image is a cut-out image of a portion where one hanger 102 is photographed.

Next, in Step P11, the SIFT feature amount is extracted from the template image, and template data is created. In addition, as indicated by black circles in FIG. 5, a plurality of SIFT feature amounts exist on the template image, and all the template data are generated. In addition, as for the method of obtaining the SIFT characteristic amount, the method described in the above Non-Patent Document 1 is used, and a description thereof will be omitted.

Next, in Step P12, the image creating section 10 creates an image photographed by the line sensor 1 as an input image. Further, as shown in Fig. 6, the vertical axis of the input image is height, and the horizontal axis is time.

Next, in step P13, the SIFT feature quantity is extracted from the input image.

Next, in Step P14, the corresponding point between the SIFT feature amount of the template data and the SIFT feature amount of the input image is searched to detect the entire hanger position in the input image. In the method of detecting the hanger position, the Euclidean distance between the SIFT feature amount of the template data and the SIFT feature amount of the input image is calculated and the corresponding point is set as the minimum value. Likewise, the Euclidean distance is determined for the entire SIFT feature amount of the template data, and the hanger position is detected based on the corresponding point.

Here, the Euclidean distance is obtained by squaring the difference between the SIFT feature amount of the template data and the SIFT feature amount of the input image and calculating the square root of the sum of them. Specifically, the SIFT feature amount is a 128-dimensional feature amount. Therefore, the Euclidean distance is calculated as shown in the following equation (1).

(1)

(One)

In the above equation (1), V is the SIFT feature quantity of the input image, V ^t is the SIFT feature quantity of the template data, and d is the Euclidean distance.

In addition, a point at which the Euclidean distance is minimized usually shows only one point in one input image. In addition, depending on the running speed of the vehicle 100, there is a possibility that a plurality of hanger 102 is photographed in one input image, and all the hanger 102 can not be detected in this state.

Incidentally, in the input image in which the horizontal axis taken by the line sensor 1 is time, a plurality of hanger 102 does not exist in the same time. Therefore, in the present embodiment, as shown by the broken line in Fig. 7, one input image is divided into short sheets in time, and corresponding points are searched so that a plurality of hanger 102 can be detected . The division width is arbitrary, but is basically the same as the template data width.

Next, in step P15, the time is calculated from the detected position of the hanger, and the time is recorded. As shown in Fig. 6, the horizontal axis represents time in the input image. Then, it is possible to calculate the time based on the sampling frequency of the line sensor 1 and the hanger position on the input image.

For example, when the sampling frequency of the line sensor 1 is 1 kHz and the hanger 102 is detected in the 100th line in the time axis direction, the photographing time per line is 1 [sec] / 1000 [Hz] = 0.001 sec] and 100 [lines] = 0.1 [sec] because the hanger 102 is detected on the 100th line.

Finally, in Step P16, Steps P12 to P16 are repeated until the process is completed for all the input images, and at the time when the process for all the input images is completed, the electric railroad repair vehicle position measurement And terminates the series of processes in the apparatus.

As described above, according to the electric railway repair repairing vehicle position measuring apparatus according to the present embodiment, the corresponding points are searched for using the SIFT feature quantity of the template data and the SIFT feature quantity of the input image, , It is possible to detect the hanger 102 even when the hanger 102 is present at the same position as the surrounding structure 104, thereby preventing the hanger 102 from being undetected.

Further, according to the electric railway repair repairing vehicle position measuring apparatus according to the present embodiment, since accurate SIFT characteristic amounts are used for scale, rotation, and brightness change, template data for all cases corresponding to scale, rotation, It is possible to detect any one template data even if there is a change in scale, rotation, or brightness. Therefore, detection efficiency of the hanger 102 can be increased.

Example 2

A second embodiment of an electric railway repair vehicle position measuring apparatus according to the present invention will be described below.

The electric railroad repair vehicle position measuring device according to the present embodiment has substantially the same configuration as that of the electric railway repair vehicle position measuring device according to the first embodiment. However, when the previous hanger position exists, But differs in that the process is speeded up by limiting the processing area as a reference.

Next, the construction of an electric railway repair vehicle position measuring apparatus according to the present embodiment will be described.

Fig. 8 is a block diagram showing a configuration of an electric railway repair vehicle position measuring apparatus according to the present embodiment.

As shown in Fig. 8, the construction of the electric railway repair vehicle position measuring apparatus according to the present embodiment is substantially the same as that of the electric railway repair vehicle position measuring apparatus according to the first embodiment, 15), when the previous hanger detection position is present, the previous hanger position and the inclination data of the hanger line (103) and the curved line (101) are received and the car which defines the processing area based on the previous hanger position And the calculation of the conference processing area is performed.

Next, the procedure of the electric railway repair vehicle position measuring apparatus according to the present embodiment will be described.

The procedure of the electric railway repair vehicle position measuring apparatus according to the present embodiment is substantially the same as that of the electric railway repair vehicle position determining apparatus according to the first embodiment, , The calculation processing of the next processing area for defining the processing area based on the previous hanger position is added.

Since the SIFT feature amount takes too much time for the calculation process, it is necessary to increase the processing speed. Since the time required for calculating the SIFT feature amount is proportional to the size of the processing region, it can be efficiently processed if the processing region can be limited. Since the hanger 102 is necessarily located between the wire 103 and the wire 101 and is more likely to be present in the inclined direction of the wire 103 and the wire 101, It is possible to define the processing region. As described above, since the processing area can be limited by the inclination of the first line 103 and the second line 101, the processing speed can be increased.

Fig. 9 is a flowchart showing the procedure of the electric railway repair vehicle position measuring apparatus according to the embodiment.

As shown in Fig. 9, in step P10, an image for a template is acquired from the line sensor 1 as a processing procedure in an electric railway repair vehicle position measuring apparatus according to the present embodiment. Further, as shown in Fig. 4, the template image is a cut-out image of a portion where one hanger 102 is photographed.

Next, in Step P11, the SIFT feature amount is extracted from the template image to generate template data. In addition, as indicated by black circles in FIG. 5, a plurality of SIFT feature amounts exist on the template image, and all the template data are generated. The method for obtaining the SIFT characteristic amount is described in the above Non-Patent Document 1, and the description thereof will be omitted.

Next, in Step P12, the image creating section 10 creates an image photographed by the line sensor 1 as an input image. Further, as shown in Fig. 6, the vertical axis of the input image is height, and the horizontal axis is time.

Next, in Step P20, if there is a previous hanger position, the next processing region calculation processing is executed in Step P21. If there is no previous hanger position, the hanger position is detected from the entire input image in step P13.

Next, in step P21, the hanger 102 is detected from the first image, and based on the inclination of the first hanger detection position 4, the hitch 103, and the crooked line 101, 5). More specifically, as shown in Fig. 10, when the hanger 103 and the hanger 101 are inclined in the lower right direction, the hanger 102 exists between the hanger 103 and the hanger 101, 103 and the torsion bar 101 in the oblique direction. Therefore, it is possible to determine the processing area 5 in the slanting direction of the tangent line 103 and the cantilever 101 with reference to the first hanger position 4.

In Fig. 10, the one-dot chain lines indicated by A and B are broken lines, the left one-dot chain line A represents the first image, the one-dot chain line A and the one-dot chain line B represents the second image, Is regarded as the third image.

Next, in step P14, the hanger position is detected in the next processing area 5 or the entire image.

Next, in step P15, the time is calculated from the detected position of the hanger, and the time is recorded.

Finally, when the previous hanger position is present in the step P16, the steps P12 to 15 are repeated until the input image disappears, and at the time when the processing is completed for all the input images, The series of processing in the repairing vehicle position measuring apparatus is terminated.

As described above, according to the electric rail repair repairing vehicle position measuring apparatus according to the present embodiment, in addition to the effect obtained by the electric rail repair repairing vehicle position measuring apparatus according to the first embodiment, (5), the processing speed can be increased.

Example 3

A third embodiment of an electric railway repair vehicle position measuring apparatus according to the present invention will be described below.

The electric rail repair repairing vehicle position measuring apparatus according to the present embodiment has substantially the same structure as that of the electric rail repair repairing vehicle position measuring apparatus according to the second embodiment. However, in the processing for limiting the processing region to the next processing region 5 And the hanger position detection accuracy is improved by expanding the next processing area 5. [

Next, the procedure of the electric railway repair vehicle position measuring apparatus according to the present embodiment will be described.

In Fig. 11, the one-dot chain line indicated by A is the place where the image is broken, the left side of the one-dot chain line A is the first image, and the right side of the one- As shown in Fig. 11, when the hanger 102 is photographed in the place where the image is broken, there is a possibility that the hanger 102 can not be detected because the SIFT characteristic amount becomes different. Thus, as shown in Fig. 11, by expanding the range of the next processing area 5 to include a part of the previous image, that is, the first image, it is possible to detect the hanger 102 photographed at the place where the image is broken .

As described above, according to the electric rail repair repairing vehicle position measuring apparatus according to the present embodiment, in addition to the effect obtained by the electric rail repair repairing vehicle position measuring apparatus according to the second embodiment, It is possible to detect the hanger 102 with good precision even when the hanger 102 is photographed at the place where the image is broken as shown in Fig.

Industrial availability

The present invention is preferably applied to an electric railway repair vehicle position measuring apparatus for measuring the position of an electric railway vehicle used for maintenance such as a curb and a track.

1: Line sensor
2: Lighting
3: Processing PC
4: Hanger position
5: Secondary processing area
10:
11, 12: Memory
13: Template creation processor
14: SIFT characteristic amount extraction processing section
15: hanger position detection processing unit
16: Time recording processor
100: vehicle
101: wire
102: Hanger
103: Hatching
104: Structure

Claims (4)

A line sensor for capturing an image of a hanging line hanging a hanging line or a hanging line supporting the hooking line,
An image creating unit that creates an image captured by the line sensor as a template image and an input image;
A storage unit that stores the template image and the input image created by the image creation unit;
A template creation processing unit for extracting SIFT feature amounts from the template image stored in the storage unit and creating template data and storing the template data in the storage unit,
An SIFT feature extraction processor for extracting a SIFT feature quantity from the input image stored in the storage and storing the SIFT feature quantity in the storage,
A hanger position detection processing unit for searching a corresponding point between the SIFT feature amount of the template data stored in the storage unit and the SIFT feature amount of the input image, detecting a hanger position in the input image, and storing the detected hanger position in the storage unit,
And a time recording processing unit for calculating the time when the hanger is detected based on the sampling frequency of the line sensor and the hanger position on the input image stored in the storage unit and storing the time data in the storage unit Vehicle position measuring device for electric railway maintenance. The method according to claim 1,
Wherein when the previous hanger position is present in the hanger position detection processing unit, the previous hanger position and the inclination data of the hanger and the crossover are received and the next processing zone And the calculation of the position of the vehicle is performed. 3. The method of claim 2,
And the hanger position detection processing section extends the second processing zone. A line sensor for capturing an image of a hanging line hanging a hanging line or a hanging line supporting the hooking line,
An image creating unit that creates an image captured by the line sensor as a template image and an input image;
A storage unit that stores the template image and the input image created by the image creation unit;
A template creation processing unit for extracting SIFT feature amounts from the template image stored in the storage unit and creating template data and storing the template data in the storage unit,
An SIFT feature extraction processor for extracting a SIFT feature quantity from the input image stored in the storage and storing the SIFT feature quantity in the storage,
And a hanger position detection processing unit for searching for a correspondence point between the SIFT feature amount of the template data stored in the storage unit and the SIFT feature amount of the input image, detecting a hanger position in the input image, and storing the detected hanger position in the storage unit,
Wherein when the previous hanger position is present in the hanger position detection processing unit, the previous hanger position and the inclination data of the hanger and the crossover are received and the next processing zone Wherein the calculation is performed and the next processing area is expanded.

Images