EP1374175A2

EP1374175A2 - Method and system for determining a change in a reference object

Info

Publication number: EP1374175A2
Application number: EP01971680A
Authority: EP
Inventors: Christian Hoffmann; Günter DOEMENS; Franz Hepp
Original assignee: Siemens AG
Current assignee: Siemens AG
Priority date: 2000-09-06
Filing date: 2001-08-29
Publication date: 2004-01-02
Also published as: NO20030992D0; NO20030992L; WO2002021445A3; NO325105B1; JP2004508643A; WO2002021445A2

Abstract

The invention relates to a method and system for determining a change in a reference object, which is associated with a travel way and which is provided at a predetermined location in the travel way. The invention provides that, at a first point in time, at least one reference image is taken at the predetermined location of the travel way. An image of the reference object is established in the at least one reference image. At a second point in time, at least one test image is taken at the predetermined location of the travel way. Another image of the reference object is established in the at least one test image. The change in the reference object is established by comparing the image of the reference object in the at least one reference image with the other image of the reference object in the at least one test image.

Description

description

Method and arrangement for determining a change in a reference object

The invention relates to the determination of a change in a reference object.

Such a determination is known from [1].

This method of image processing, a method for coding and decoding a digital image according to an image coding standard MPEG4 [9], is based on a principle of object-based image coding.

In object-based image coding, segmentation of two successive images into image blocks or image areas corresponding to objects occurring in a scene is carried out and these objects are encoded separately.

With this object-based image coding, an image of an object is determined in a so-called inter-image coding with motion estimation in a temporally previous image. Then a further image of the object is also determined in the image that follows in time. A difference image information is determined using image information associated with the image of the object in the temporally preceding image and the image information associated with the further image of the object in the temporally subsequent image. This difference image information is subsequently encoded and transmitted.

The difference image information is only very small if a change in the object in the temporally successive digital images, for example due to a movement or a change in shape of the object, is small. If the change is large, there is a lot of difference information. Out For this reason, as is known from [1], an “image-to-image * movement (motion estimation) is measured and compensated for before the difference information is determined (motion compensation).

From [2] a method of image processing for monitoring a route of a train is known.

In this method, video images of a contact wire, which is used to supply power to a train and which is guided along the travel path using a so-called longitudinal chain mechanism and a so-called transverse support device, are recorded.

In FIG. 2 and FIG. 5, a structure of an overhead line for the power supply comprising the contact wire 201, the longitudinal chain train 202 and the transverse support device 203 is shown schematically.

2 and 5 also show two insulators, a hanger 206, a support rope twist clamp 503, a tip anchor 504, a cantilever tube 505, a diagonal tube 506.

In the method known from [2], the video images are used by a specialist in terms of damage to the

Contact wire or a foreign body in the overhead line is visually checked and evaluated.

Further automatic methods for monitoring a route of a train are known from [3], [4], [5] and [6] or are proposed in [8].

For example, a triangulation method [3] and time-of-flight-based phase measurement methods [4] are used to determine the vertical and lateral position of the contact wire. To determine the wear of the contact wire, either a width of a grinding surface is measured with the aid of a laser scanner [5] or diode rows [2] or a residual height of the contact wire is determined in horizontally oriented transmitted light [6].

A measuring device used for this purpose and known from [6] is located above a grinder, which limits the vehicle speed for mechanical reasons, leads to shading in the overlap area of several contact wires and thus possibly simulates a larger residual height than actually present. To prevent this, such an area is excluded from a measurement.

A non-contact, recording-based monitoring method with an almost vertical viewing direction of an acquisition unit also has the disadvantage that only the width of a mirror surface can be detected. In principle, however, the mirror surface does not allow the residual height to be determined for a contact wire with an almost rectangular profile.

Other components of the power supply, in particular the long chain mechanism and the transverse support device, cannot be automatically inspected in the monitoring methods mentioned above. It is therefore common to include the

Power supply with a video camera, whereby its relatively low resolution and frame rate reveals only very gross defects. After an inspection trip, all video images must be visually checked and evaluated by a specialist or inspection staff.

Alternatively, a route is walked by the test personnel, which is very slow and therefore costly and also involves a certain risk of accident.

The invention is therefore based on the problem of specifying a method and an arrangement with which automatic and on a change in an object in a route is ascertained in a simple manner and the route can thus be monitored in a simpler and more cost-effective manner and with greater reliability than in the methods mentioned above.

Brief description of the invention

The problems are solved by an arrangement and a method with the features according to the respective independent patent claim.

The following steps are carried out in the method for determining a change in a reference object which is associated with a route and which is present at a predetermined location in the route:

at least one reference image is recorded at the predetermined location of the route,

an image of the reference object is determined in the at least one reference image, at least one test image is recorded at the specified location of the route at a second point in time,

a further image of the reference object is determined in the at least one test image,

The change in the reference object is determined by comparing the image of the reference object in the at least one reference image with the further image of the reference object in the at least one test image.

The arrangement for determining a change in a reference object which is associated with a route and which is present at a predetermined location in the route has the following components:

a first recording unit, which is set up in such a way that at least one reference image can be recorded at the predetermined location of the route, a first evaluation unit, which is set up in such a way that an image of the reference object can be determined in the at least one reference image,

a second recording unit, which is set up in such a way that at least one test image can be recorded at the predetermined location of the route,

a second evaluation unit that is set up in such a way that a further image of the reference object can be determined in the at least one test image, a third evaluation unit that is set up in such a way that by comparing the image of the reference object in the at least one reference image with the change of the reference object can be determined in the further image of the reference object in the at least one test image.

The arrangement is particularly suitable for carrying out the method according to the invention or one of its further developments explained below.

Preferred developments of the invention result from the dependent claims.

The further developments described below relate both to the method and to the arrangement.

The invention and the further developments described below can be implemented both in software and in hardware, for example using a special electrical circuit.

Furthermore, an implementation of the invention or a further development described below is possible by means of a computer-readable storage medium on which a computer program which carries out the invention or further development is stored. The invention or any further development described below can also be implemented by a computer program product which has a storage medium on which a computer program which carries out the invention or further development is stored.

Reliability in determining the change in the reference object can be increased by taking several reference images and / or several test images at the specified location. It is expedient to record the reference images using different recording parameters, for example a recording angle. The same applies to the test images.

The reliability can be further increased if the reference image and the associated test image are recorded under the same recording parameters.

To reduce storage space and computing effort, it is expedient to determine a first partial image in the reference image and / or a second partial image in the test image, which have the image or the further image. In this case, the first partial image and the second partial image are compared with one another in the comparison.

It may also be useful to place the first partial image in the reference image and / or the second partial image in the test image such that the first partial image and / or the second partial image shows an environment associated with the reference object.

In one embodiment, a partial image is a search mask. Using this search mask, a defined image area can be examined for the image, the further image or the change.

The comparison of the first with the second partial image can be realized particularly simply by the fact that a first ter pixel value of a first pixel of the first field is subtracted from a second pixel value of a second pixel of the second field. Such a first pixel value and such a second pixel value can be, for example, brightness or color values (luminance or crominance values).

This subtraction of pixel values produces a difference image in which a change in the image between the image and the further image can be determined particularly easily. The change in the reference object can be determined using the image change.

The difference image can also be processed further, for example by filtering and / or segmenting an image area of the difference image. This can increase the reliability in determining the change.

In one development, the image of the reference object in the reference image or the further image of the reference object is determined using an edge detection method or object recognition method. Such methods are known from [10] and [11].

The further image of the reference object in the test image can be determined in a particularly simple manner using a correlation method as is known from [9], [12] and [13]. The correlation between pixel values which belong to the image and pixel values which belong to the further image is determined.

The reliability in determining the change in the reference object can also be increased in that a first quantity of a geometric reference property of the reference object is determined using the image of the reference object in the reference image and one using the further image of the reference object in the test image second size of the same geometric reference property of the reference object is determined.

As the geometric reference property, for example, an angle which the reference object has, or a length which the reference object has, or a curvature which the reference object has, or a distance which the reference object has, or a shape which the reference object has, be determined.

In this case, the comparison compares the first size with the second size. In this comparison, a difference can be formed between the first and the second quantity and the change can be determined when the difference exceeds a threshold value.

In one embodiment, the reference image and the test image are assigned to one another using a location code and / or a time code. Such a location code can be, for example, a number which is assigned to the reference image and / or the test image. Such a time code can, for example, be a time specification which is assigned to the reference image and / or the test image.

In one embodiment, several locations are specified. The method according to the invention is carried out at each of the predetermined locations.

In this case, appropriate further training can be used to monitor the route. Changes to reference objects can be determined at all specified locations and a measure can be taken to remedy the corresponding change.

In a further development, the reference object is broken down into several reference partial objects. In one configuration, the reference object is a functional accessory for the route, for example a traffic engineering object. In this case, the route can be a traffic connection, for example a railway line.

In a further embodiment, the reference object is an element of a contact line. Such an overhead line can be an overhead line. In this case, the reference object is an element of a longitudinal catenary of the overhead line or a transverse support device of the overhead line.

A reference object or one of its reference sub-objects can also be a basic element of the overhead contact line, in particular a rod or a tube or a rope or a wire, or a connecting element of the overhead contact line, in particular a holder or a clamp or a carrying element or an insulator.

In a further development, the reference object is an element of a rail track.

A particularly simple arrangement of the invention can be realized in that the first recording unit and / or the second recording unit is in each case a digital camera. An evaluation unit can be implemented in a particularly simple manner by means of a PC on which a computer program is stored, with which corresponding method steps can be carried out.

In a further development which is used to monitor the route, the arrangement is attached to a vehicle which is traveling on the route.

An embodiment of the invention is shown in figures and is explained below. Brief description of the figures

Show it

FIG. 1 shows a principle of an image evaluation by recognizing a change in an overhead line object;

Figure 2 is a sketch of an overhead line of a railroad;

Figure 3 shows an inspection vehicle according to the embodiment;

Figure 4 is a sketch of a recording system for receiving a longitudinal chain;

FIG. 5 shows a field of view for an inspection of a transverse support device;

FIG. 6 shows a sketch which shows a determination of an evaluation window;

Figures 7a and 7b sketches, which illustrate methods for inspecting an overhead line according to an embodiment;

Figure 8 is a sketch illustrating a method for inspecting an overhead line according to an embodiment;

Figure 9 is a sketch with images that illustrate a definition of a search mask.

Exemplary embodiment: automated optical inspection of an overhead line of a railway connection

System concept of automated optical inspection Fig. 3 shows an inspection vehicle 300, a modified railcar, for monitoring a railway line.

The aim of the inspection is to detect a change in an overhead line of the railway line.

Such a change can arise, for example, from damage that occurs to an element of the overhead line or from a foreign body that gets caught in the overhead line.

If a change is detected and classified as essential for the trouble-free operation of the railway line, a measure to remedy the change is initiated. Such a measure can be, for example, repairing the damage or removing the foreign body.

It is pointed out that the inspection system described below can be used not only for the inspection of an overhead line of a railway connection, but with a corresponding modification for the inspection of any element of a railway connection, for example a rail body or a rail sleeper.

On a roof 301 of the inspection vehicle 300 there are a plurality of optical recording systems 302, 303, 304, 305 which, during an inspection trip of the inspection vehicle 300, record contact-free images of the overhead line from different directions at given recording locations 300 from different directions.

The optical recording systems 302, 303, 304, 305 are aligned in such a way that all the elements of the overhead line as well as their surfaces can be recognized on the pictures recorded by them, each belonging to one another for a location. The optical recording systems 302 and 303 are used for monitoring and inspection of the transverse support devices 306, the optical recording system 304 is used for monitoring and inspection of the longitudinal catenary 307 and the optical recording system 305 is used for monitoring and inspection of the contact wire 308.

2 shows an example of a recording location with an overhead line and elements of the overhead line, for example a longitudinal catenary 310 (see FIG. 2) and a transverse support device 311 (see FIG. 2).

All captured images are stored in a location and time-coded manner, so that each image can be uniquely assigned an associated location and time.

At the end of an inspection trip, the stored images are evaluated with the help of an automatic image processing method described below.

As part of this evaluation, the images of the inspection trip are compared with reference images taken at an earlier point in time at the same predetermined locations under the same recording parameters. A reference image is assigned to an associated inspection image using the location and time codes assigned to the images.

In this comparison, a change in the overhead line or an element of the overhead line is determined, this is recorded in a log and automatically classified into one of three classes: clear defects, unclear situations and insignificant differences.

Only unclear situations are subsequently assessed by test personnel. Finally, a maintenance notice or a repair measure is initiated if a change is found.

Since the test personnel only need to visually inspect the unclear situations, they are freed from the time-consuming and tiring visual inspection of the much larger, flawless areas.

In order to be able to recognize smaller elements of the overhead line such as a screw head and individual wires protruding from defective ropes, the resolution of the recording systems 302, 303, 304, 305 must be approximately 1 to 2 mm. In order to avoid greater motion blur in an image caused by a vehicle speed of 80 km / h, the exposure time of a recording system 302, 303, 304, 305 must not be longer than approximately 45 microseconds. This results in a data volume to be saved of 100 MByte / s or 5 GByte / km.

A more precise specification of a recording system will be made at a later date.

When the contact wire 308 is measured for wear, a comparison of inspection images with earlier reference images is not necessary. In this case, it is sufficient to compare a profile of the contact wire 308 recorded during the inspection trip with a predetermined target profile. A measurement uncertainty of the corresponding recording system 305 must be approx. 0.1 mm and enable the detection of a profile constriction with a length of 2 to 3 cm.

Since the duration of a single measurement for the longitudinal chain 307 does not exceed 45 microseconds or for the wire profile measurement 200 microsec, a vibration of the inspection vehicle 300 or of the contact wire 308 with a frequency below 5000 Hz has no influence on the quality of a measurement result. For a larger frequency, an amplitude is smaller than a required resolution and therefore does not have a disturbing effect.

Automated optical inspection system

The arrangement shown in FIG. 4 serves to accommodate the longitudinal chain mechanism 307 or 400. On the left side 401 and on the right side 402 of a vehicle roof 403 there are each a diode line camera 404, 405 with associated lighting 406, 407.

Each diode line camera 404, 405 records a vertical image strip on one side of the longitudinal chain mechanism 400 within 45 microseconds. In connection with a horizontal vehicle movement of the inspection vehicle 300, a

"infinitely" extended picture. 22,000 lines / s are recorded at a vehicle speed of 80 km / h, each strip being 1 mm wide. The camera line contains 2048 elements, which corresponds to a resolution of 0.7 to 2 mm in the direction of the camera line, depending on the distance.

In order to enable use at night and to create defined lighting conditions during the day, the longitudinal chain 400 is illuminated. Metal vapor headlights 406, 407 with a connected load of approximately 10 kW were initially considered as lighting 406, 407. However, since space problems arose due to a large number of cameras and glare from an uninvolved party was to be avoided, an infrared diode laser was used. In contrast to a metal vapor headlight, this illuminates exactly one field of view of the respective camera and no additional areas of no interest, so a power of 15 W is sufficient.

In order to rule out any danger to a human eye, the infra- U ⁾ cυ tv> MP ¹ t- ¹ cπ o cπ O Cπ o cπ

C_

3 y.

3

•

Any image disturbance that may occur, for example due to a swaying of the inspection vehicle, can be compensated for using continuously logged data from tank sensors.

A laser radar system 322, 323 [6] is located on the roof 301 near both end faces 320, 321 of the inspection vehicle 300, the beams of which radiate vertically upward recognize a contact wire side holder of a transverse support device 306.

As soon as the laser radar system 322 in front in the direction of travel detects a transverse support device 306, it triggers an image recording by a scanner camera 324 oriented in the direction of travel. After passing the transverse support device 306, a rear-facing scanner camera 325 is triggered. This creates pictures of an object from both sides.

3, fan beams 331, 332, 333 of the recording systems 302, 303, 304 are indicated. From radiation fans for that

Longitudinal chain mechanism 307 is drawn for the sake of clarity, only one beam fan 333. This is approximately vertical on a plane of the longitudinal chain mechanism 307, while the beam fans 331, 332 for the transverse support devices 306 are arranged almost horizontally. Environmental influences that affect the measurements are minimized by the following measures:

In order to be able to measure during the day as well as during the night, narrow-band filters in front of the cameras only allow light with the wavelength of the infrared lasers described to pass through. This means that sunlight, as long as it does not shine directly into the cameras, can be suppressed.

Because the temperatures on the roof can fluctuate more than the lasers, cameras and optical systems can handle the inside of all housings is tempered to about 20 C using closed water circuits.

The windows in front of the optics are safely protected against light rain during the measurement. In the case of very heavy rain, fog or snowfall, the view of the transverse support devices more than 10 m away can be obstructed, so that u. U. good measurement results can no longer be achieved.

A cleaning of the windows from dirt, such as. B.

Dust and soot can be carried out automatically with a warm washing liquid that is sprayed on with pressure. The vehicle roof does not need to be stepped on here. If no measurements are taken, protective covers close the windows.

image analysis

A principle of an image evaluation 100 in the automated optical inspection of an overhead line of a railway connection is shown in FIG. 1 and is explained in the following.

The images (inspection images) 101 recorded during an inspection trip are transmitted to computers (industrial PCs) which are located inside the vehicle of the inspection vehicle. First, the inspection images 102 are compressed 103 by means of digital signal processors to approximately 3% of their original data volume and, after being temporarily stored on a hard disk, transferred to a magnetic tape.

Such a magnetic tape can store all data from around 300 km of single track. Because of the hard disk acting as a buffer, it is not necessary to interrupt the measurement when changing the tape.

The automatic evaluation 110 can be carried out after the inspection trip, during longer waiting breaks or trips without Measurement. The start of the evaluation 110 on the vehicle can be brought forward by installing additional computing technology.

A prerequisite for the image evaluation 110 is that a so-called reference run was carried out at an earlier point in time, during which the actual state (reference state) was recorded, assessed and reference data generated using reference images 102.

The reference images 102 and the inspection images 101 each correspond in terms of their location and in their exposure parameters.

FIG. 1 further shows a comparison 120 of a reference image 102 from the reference run with the corresponding inspection image 101 from a current inspection run.

As part of the image evaluation 110 or 130, three methods are carried out for the detection of changes in the overhead line, which are described below for the detection of changes in the transverse support devices and in a longitudinal chain mechanism:

a) method based on a geometry comparison 131, b) method based on a difference image 132, c) method for recognizing a foreign body 133.

Basic principle of procedures a) to c]

Each of these three methods 131, 132, 133 is based on a comparison 130 of a reference image 102 with the corresponding test image 101. In this comparison 130 and 135, an image 136 of a selected object is first determined in the reference image. An image 137 of this selected object is then likewise determined in the corresponding test image. A change 135 of the selected object is determined 140 by comparing 135 the two images 136, 137 of the selected object or by comparing measured values which are determined using the images

Regarding methods a) and b):

6 shows the type and structure of reference data for longitudinal chain mechanisms 601, 602. They are derived from a large number of very large overall images, each of which comprises the longitudinal chain mechanism 601, 602 between two transverse support devices 603, 604. The reference data contain small drawing files 611, 612, 613, 614, the components such. B. terminals 621, 622 in the target state, their geometric dimensions and relations and route-related position information. Ropes 631 and contact wires 632 are defined by their diameter and the fastening points 633, 634 located in the partial images.

With the help of the automatic detection of the transverse support devices 603, 604 and the continuous logging of the pulses of an odometer, a subsequent assignment of the recording location is guaranteed to be more precise than 0.5 m. In the automatic comparison, because of the existing knowledge of the recording locations, the related images of the reference and inspection trip can easily be found by the computer.

FIG. 7a schematically shows a reference image 701 of a partial area of a transverse support device 702 (without contact wire and suspension cable). Sub-images or evaluation windows 703, 704, 705, 706, 707 are identified by bright frames.

7b shows the corresponding test image 710. However, a terminal 720 (in the reference image) or

721 (in the test image) moved and a protruding rope end 722 inserted as a defect. Image processing finds the existing changes and outputs them as a message on the defect report.

Moving the clamp 720 or 721 has resulted in changes in the distance 724 and angle 725. The protruding rope end 722 is clearly visible. Other changes to other components are found in a similar manner.

Regarding procedure c):

8 shows a partial area of a transverse support device 801 in a test image 800.

The search image shows two search masks 802, 803 which were previously defined in the corresponding reference image.

These search masks are generally defined in a reference image in such a way that they cover image areas, so-called search areas, between displayed objects. A foreign body can thus be identified when a search area in the reference image is compared with the corresponding search area in the test image.

This procedure clearly shows that a search area from the reference image (possibly after an adjustment to a changed position of the objects in the test image) is used in the test image by searching for foreign objects within the adapted search area in the test image, which are characterized by their brightness from i.allg. stand out against dark background.

8 shows a loose rope end 810, which is recognized when the search area 802 of the test image 800 is compared with the corresponding search area in the associated reference image, in which the loose rope end is not present. In this way, an additional object, for example a foreign body, which is depicted in the test image but not depicted in the reference image, can be recognized in a simple manner.

A search mask is defined according to the following procedure (Fig. 9):

First, a selected reference image 900, which has different objects 901, 902, 903 and whose pixels have different gray values 904, 805, is binarized 810.

In the binarized reference image 911, the objects 901, 902, 903 appear white. The remaining areas appear black.

In a further step 920, the areas shown in white are widened (Figure 921).

The search areas 932, 933, 934 are defined in such a way that each associated black area is surrounded by a search mask. The different parts of the search mask are combined and saved in a single binary search mask image.

An alternative to the exemplary embodiment is given below:

This alternative describes a method for location coding for the assignment of reference image to test image:

So that the corresponding reference image of the same location can be assigned to a test image, location information from a position measuring system is continuously recorded for the images, for the reference images as well as for the test images during the respective recording journey. In addition, a special system for recognizing a main object (e.g. the cross support device for electrical railways) the presence of a cross support device and thus a support mast with the associated location is also recorded.

In this way, the crossbeams and thus their masts can be counted and numbered along the route. During inspection, these numbers are used to assign the correct reference images to the corresponding test images.

The following publications are cited in this document:

[1] ISO / IEC JTC1 / SC29 / WG11, MPEG-4 Video Verification Model Version 5.0 Doc. N1469, Nov. 1996, pp. 55-59;

[2] J.M. Van Gigch, C. Smorenburg, A.W. Benshop, "System for Measuring the Contact Wire Thickness (ATON) of the Dutch Railways", Schienen der Welt, April 1991, pp. 20-31

[3] B. Sarnes, "Quality Assurance for Pantographs and Transmissions", ETR, 48 (1999) 3, pp. 117-123

[4] H. Höfler, N. Dimopoulos, B. Metzger, "Contact wire position measurement on high-speed lines", eb, 96 (1998) 3 pp. 61-64

[5] S. Ishizu, "Series 700 derivative will replace Doctor

Yellow ", Railway Gazette Intern., April 1999, pp. 218-220

[6] P. Pohl "The Hagen video measurement railcar *,

DerEisenbahningenieur, June and July 1996, booklets 6 and 7 1996, pages 24 and 32 respectively

[7] R. Müller, H. Höfler, "Track Monitoring with Optical Measurement Technology", Railway Engineer Calendar '97, Darmstadt: Tetzlaff, 1996, pp. 315-332

[8] DE 199 36 448 AI

[9] available on September 1, 2000 at http: //www.it.kth. s / docs / iso 13818-2.read.html

[10] available on September 1st, 2000 at: http: // vyll .mit. fh-heilbronn.de/German/Theorie/Kantendet ektion.htm [11] available on September 1, 2000 at: http: //www.ifi .unizh. ch / groups / mml / people / mbichsel / lecture .html

[12] B. Jahne, Digitale Bildverarbeitung, Springer Verl. 1997

[13] F. M. Wahl, digital image signal processing, chapter 6.5, detection filter, Springer Verl. 1984.

Claims

claims

1. A method for determining a change in a reference object which is associated with a route and which is present at a predetermined location in the route, with the following steps:

an image of the reference object is determined in the at least one reference image,

at least one test image is taken at the predetermined location on the route,

a further image of the reference object is determined in the at least one test image, the change in the reference object is determined by comparing the image of the reference object in the at least one reference image with the further image of the reference object in the at least one test image.

2. The method according to claim 1, in which a plurality of reference images are recorded at the predetermined location.

3. The method according to claim 1 or 2, in which a plurality of test images are recorded at the predetermined location.

4. The method according to any one of claims 1 to 3, in which a first partial image is determined in the reference image, which has the image.

5. The method according to any one of claims 1 to 4, in which a second partial image is determined in the test image which contains the further image.

6. The method according to claim 5, in which the first partial image and the second partial image are compared with one another in the comparison.

7. The method according to claim 6, in which a difference image is determined in the comparison using the first partial image and the second partial image by subtracting pixel values of one image point of the first partial image and one image point of the second partial image.

8. The method according to claim 7, in which an image change between the image and the further image is determined in the difference image.

9. The method according to claim 8, in which the change in the reference object is determined using the image change.

10. The method according to any one of claims 1 to 9, in which in the using the image of the reference object in the reference image a first size of a geometric reference property of the reference object is determined and using the further image of the reference object in the test image a second size of the same geometri - See reference property of the reference object is determined.

11. The method according to claim 10, in which, as the geometric reference property, an angle which the reference object has, or a length which the reference object has, or a curvature which the reference object has, or a distance which the reference object has, or a Shape that the reference object has is determined.

12. The method of claim 10 or 11, wherein the comparison compares the first size with the second size.

13. The method according to / claim 12, in which a difference is formed between the first and the second variable and the change is determined when the difference exceeds a threshold value.

14. The method according to any one of claims 1 to 13, wherein the reference image and the test image are assigned to one another using a location code and / or a time code.

15. The method as claimed in one of claims 1 to 14, in which a plurality of the locations are specified and the method according to one of the preceding claims is carried out at each of the specified locations.

16. The method according to any one of claims 1 to 15, used to monitor the route in such a way that the change is determined and a measure for eliminating the change is carried out.

17. Arrangement for determining a change in a reference object which is associated with a route and which is present at a predetermined location in the route, with the following components:

a first recording unit, which is set up in such a way that at least one reference image can be recorded at the predetermined location of the route,

- A first evaluation unit, which is set up in such a way that an image of the

Reference object can be determined,

a second recording unit, which is set up in such a way that at least one test image can be recorded at the specified location of the route at a second point in time, a second evaluation unit that is set up in such a way that a further image of the reference object can be determined in the at least one test image is a third evaluation unit, which is set up such that the change in the reference object can be determined by comparing the image of the reference object in the at least one reference image with the further image of the reference object in the at least one test image.

18th Arrangement according to claim 17, in which the reference object comprises a plurality of reference partial objects.

19. The arrangement according to claim 17 or 18, wherein the reference object is a functional accessory of the route.

20. Arrangement according to one of claims 17 to 19, wherein the reference object is a traffic engineering object.

21. Arrangement according to one of claims 17 to 20, in which the route is a traffic connection.

22. Arrangement according to one of claims 17 to 21, wherein the traffic connection is a railway line.

23. Arrangement according to one of claims 17 to 22, wherein the reference object is an element of a catenary.

24. The arrangement according to claim 23, wherein the catenary is an overhead line.

25. The arrangement as claimed in claim 24, in which the reference object is an element of a longitudinal chain of the overhead line or a transverse support device of the overhead line.

26. Arrangement according to one of claims 18 to 25, in which a reference subobject of the plurality of reference subobjects is a basic element of an overhead contact line, in particular a rod or a tube or a rope or a wire, or a connecting element of the overhead contact line, in particular a holder or a clamp or a carrying element or an isolator.

27. / Arrangement according to one of claims 17 to 22, wherein the reference object is an element of a rail track.

28. Arrangement according to one of claims 17 to 27, wherein the first recording unit and / or the second recording unit is in each case a digital camera.

29. Arrangement according to one of claims 17 to 28, used for monitoring the route, the arrangement being attached to a vehicle which is traveling on the route, the change can be determined and a measure for eliminating the change can be carried out.