WO2013167255A2

WO2013167255A2 - Method, signal sequence and computer installation for creating, managing, compressing and evaluating 3d data from a three-dimensional terrain model, and a computer program with program code for carrying out the method on a computer

Info

Publication number: WO2013167255A2
Application number: PCT/EP2013/001317
Authority: WO
Inventors: Konrad Winkler; Andreas Hartwig
Original assignee: Db Systel Gmbh
Priority date: 2012-05-09
Filing date: 2013-05-04
Publication date: 2013-11-14
Also published as: DE102012009688A1; WO2013167255A3; DE102012009688B4

Abstract

The invention relates to a method, a data storage medium having data stored thereon or a signal sequence that is suitable for sending via a data network and that represents data, and also to a computer installation for creating, managing, compressing and evaluating 3D data from a three-dimensional terrain model and a computer program with program code that performs the method steps when the computer program is executed on a computer or a computer network. The volume of data is produced in the database by varying the position and time coordinates of the data points from the 3D environment and the objects situated therein on the basis of the relevance of the objects. Depending on the performance of the display unit, the data are forwarded to the terminal such that the point density in the presentation is matched to the performance of the display unit or of the terminal. In this case, the data are presented on an output unit such that object categories with low relevance are presented at a lower resolution than object categories with higher relevance.

Description

Method, signal sequence and computer system for creating, managing, compressing and evaluating 3D data of a three-dimensional terrain model and a computer program with program code for performing the method on a computer

The invention relates to a method, a data carrier with data stored thereon, or a signal sequence representing data suitable for transmission over a data network and a computer system for creating, managing, compressing and evaluating 3D data of a three-dimensional field model and a computer program Program code that performs the method steps when the computer program is executed on a computer or a computer network.

Using 3D models, an existing environment in the real world can be created and examined in a data processing device. By visualizing this environment on a display device, users can perform valuable analyzes in the environment without having to investigate on-site, possibly under adverse circumstances. In the data model, it is possible to simulate processes that in reality can not be investigated at all or only with great effort. Furthermore, the 3D model can be specifically changed in the data processing system to illustrate plans.

The advantages of such three-dimensional digital terrain models have been known for some time. Different ways were developed to obtain, evaluate and manage the required terrain data.

DE 10 2009 006 085 A1 describes a method for acquiring route data for route networks of track-guided vehicles. In this case, position data of a route are first acquired by means of satellite-aided generated data. The acquired position data are assigned to a raster point, wherein a value for the geographic height of the raster point is determined for each of the raster points from a height database. Then successive position data are linked to their respective height and combined to form a route. In addition, data from the terrestrially measured markings for the route kilometers and recordings can be obtained the route can be compared or supplemented by the one or more cameras with the other data.

A particular disadvantage of this method is that no data reduction or data compression takes place. The data is simply adapted to a fixed grid, which ultimately defines the accuracy and thus the amount of data for the entire model. All data is matched to this rigid grid to compare.

DE 10 2007 045 082 A1 describes an apparatus and a method for updating map data for a predetermined area. Here, 3D location information is collected from a variety of mobile location devices and compared to existing map data for the area. The deviations between map data and measurement data, i. missing or contradictory sections are identified.

A major disadvantage of this method is, in particular, that a large number of uncontrolled and therefore random individual trajectories must be evaluated, whereby different reliabilities for the measured trajectories are determined in order to obtain reliable data. The object of the present invention is to provide a method with which it is possible to combine a 3D model of an environment from data of different origin or measuring methods and different accuracy into a common database, thereby reducing and reducing the amount of data of a 3D terrain model to control that different end devices of different performance can use the data for their applications in near real time. Here also movable objects in the data model are to be detected and their movement in the environment can be simulated.

It is also an object of the invention to provide a data carrier with data stored on it or for the transmission via a data network, signal representing signal sequence, which brings together a 3D model of an environment of data from different sources or measurement methods and different accuracy to a common database and In doing so, the data volume of a 3D terrain model is reduced and controlled in such a way that different end devices of different performance can use the data for their applications in near real time. Furthermore, a computer system is to be provided which is set up in such a way that it merges a 3D model of an environment from data of different origin or measuring methods and different accuracy into a common database, thus reducing and reducing the data volume of an SD terrain model controls that different end devices of different performance can use the data for their applications in near real time. In the process, moving objects should also be recorded in the data model and their movement in the environment simulated.

Furthermore, a computer program with program code is provided, which performs the method steps when the computer program is executed on a computer or a computer network. The program code may either be stored on a data carrier containing the method as a machine-readable and executable program, or in a computer network such as e.g. the Internet is made available to client-server systems running the program.

These objects are by the inventive method according to claim 1, a data carrier with stored data or suitable for transmission over a data network, signal representing signal sequence according to claim 7, a computer system, set up as a client-server system according to claim 13 and a computer program Program code solved according to claim 19.

Advantageous developments are subject matters of the dependent claims.

If a 3D data model is to be created for an environment, measurement data or design data for the environment and the surrounding objects are usually already available, which can be merged, checked and possibly supplemented by further measurements. A data point is defined by 3 spatial coordinates and optionally by a temporal coordinate. A data point can originate as a measurement point from a 3D measurement or as 3D or 4D coordinates by calculations during planning or by manual input become. Since these are generally different data obtained with different accuracies, the data must be prepared in order to create a general database.

For the performance of any application that accesses the database, care should be taken to reduce the amount of data in the database as much as possible so that only the essential data is available. In particular, when using different applications for the data, which may also be processed and displayed on different powerful terminals, the amount of data from the database needs to be further compressed in order to respect the performance of the terminals or the display units connected to them to take.

The inventive method according to claim 1 reduces the amount of data in the database by the position and time coordinates of the data points of the 3D environment and the objects therein are varied depending on the relevance of the objects.

Basically, the 3D data is present in the accuracies provided by the different data sources. The accuracy of the measured data is essentially determined by the density of the measuring points. The higher the spatial measurement point density, the more accurate the information about an object, provided that the accuracy of the individual measurement has been selected appropriately for a measurement point.

However, it is generally possible that the accuracies provided by the data sources are better, at least for subregions of the environment or for a part of the objects, than is necessary for the intended applications. For example, for a traffic simulation application, it is not necessary to treat objects outside the traffic routes with the same high accuracy as the traffic routes themselves.

Therefore, according to the invention, the objects and the environment are classified into categories (hereinafter always referred to collectively as object categories). The object categories are each assigned a minimum requirement for accuracy. The requirement for accuracy is selected by the relevance of the object category for the intended applications. If z. If, for example, all intended applications evaluate a particular object category as less relevant, the spatial density of the measured data for these object categories can be selected to be correspondingly low. Object categories, which have a assigned a high relevance, get a correspondingly higher demand on the spatial density of the data points.

Object categories of the same accuracy are assigned to corresponding clusters.

If necessary, attributes for the existing measurement data can be evaluated for classification into the object categories. Thus, for example, the measurement data, which originate from surveying, usually provided with short descriptions of the measuring points that can be used for this evaluation. For example, all objects which are to be regarded as limiting during the movement of a vehicle along a traffic route can be stored in the database with the same accuracy.

Since only a high data point density is maintained for the object categories whose treatment requires a high degree of accuracy, while other object categories are treated with a lower data point density, the data volume and hence the storage requirements for the database are significantly reduced. Furthermore, less data has to be considered for the processing of the data, which significantly increases the processing speed.

Since access to the database different applications, the data are also processed according to the invention so that users with simple devices receive the data in another, tailored to the performance of his terminal form, as users who have more powerful computers available. For example, a user with a simple navigation device is an employee with a GSM-R terminal that is to be sent to a specific location for maintenance or landscaping. The amount of data and the detail of the position determinations for the objects must here be significantly less than e.g. for a planner, who is to use the database to plan a traffic route with the associated components on a computer. Therefore, e.g. Buildings that are located on a traffic route, presented for the landscape care application with a lower resolution than for the planning application. Accordingly, a sufficient number of object categories must be created, which can be assigned different accuracies depending on the application.

The method according to the invention thus makes it possible for terminals with low performance to be able to process and display the data in a usable form. The dependent claims 2 to 4 describe advantageous embodiments of claim 1, wherein the amount of data is further reduced in each case. According to claim 2, the redundant data points are determined and deleted when merging the data from different data sources. Redundant data exists when data points have the same position data within the measurement accuracy and these data points are assigned to the same object. Redundant data is deleted so that each data point is contained only once in the database.

Furthermore, it may be advantageous to also remove valid data points from the database, which can be calculated by an interpolation method. This means that data points that are actually appropriate for the required accuracy of the assigned object category are also removed during the preparation of the data. The prerequisite for this is that this data can be calculated within the required accuracy from the surrounding data points. For line and area entities, linear interpolation methods are used, such as linear interpolation. Polylines and triangulations. For curved lines, for example, fitted curve and surface splines are used.

Advantageously, frequently occurring objects are replaced by algorithmic objects. If a data analysis shows that the measured values accurately reflect such an object within the scope of the measuring accuracy, it is not necessary to map the object by concrete measuring points. Instead, the object can be simulated or computed in a simplified form by an algorithmic object, which significantly reduces the amount of data in the database. Furthermore, it is advantageous if the measurement data are released from the noise component. Especially with 3D scans, there are certain measured data shares that can not be assigned to actually existing objects. These measurement points are filtered out by comparing the scan data with the other data in the 3D model. Deviations from the other data collected in the data model can occur when objects have been added or removed. Furthermore, objec- te have changed, for example, when a curb of a platform edge has moved. Such data is detected using object and edge detection. Data points that can not be meaningfully assigned to such analysis methods are either to be examined more closely by the processor or can be classified as noise and deleted. A closer examination is particularly indicated in the areas of the 3D model, where high security requirements are made. For this purpose, the potential noise points are stored and presented to the user with a corresponding note on the display unit.

In accordance with claim 3, weightings are also assigned to the data points in addition to measurement and display tolerances. The respective weights and their combination in a common function characterize the validity of a measure of a particular measurement method in the 3D model. When interpolating points or areas between the data points used for the data model, basis vectors of the underlying local coordinate system are used. These relate i.a. on points of different accuracy. The position vector to an interpolated point is represented as a linear combination of the basis vectors, which are weighted with the accuracy of the associated base points. This results in a linear combination of the weights, which provides information about the accuracy of the interpolated point or the interpolated surface.

Measurements between measured points have a high weighting, dimensions between measured points and interpolated points have a lower weighting.

According to claim 4, a 3D representation of a track of a railroad tracked traffic is an application field for the invention, e.g. a railway line. Here it is advantageous to define the spatial arrangement of an environment around the track by determining the distance of the objects to the track.

As a measure of the relevance of the object data safety aspects are selected.

The tracks of railway networks were generally surveyed by surveyors to the millimeter. These measurements are therefore an ideal basis for to serve as a reference system for determining the position of objects in the environment relative to the track.

The elements of the control and safety technology are i.a. measured at a lower resolution than the rails.

The same applies to structural objects that are close to the clearance gauge.

More distant objects or vegetation data can be treated with even lower resolution, as they do not exert much influence on the safety of the railway operations.

The invention also makes it possible, due to the different object categories, to use a generalized, hierarchical measuring principle according to claim 5.

In a conventional 3D model, the distance between two points is used to define the spatial order.

In the case of the hierarchical measuring principle, for example, within the model, the spatial arrangement is measured by distances between objects of the same relevance cluster. Only measurement relations are allowed which relate to matching objects, e.g. Platform and track, track and level crossing, buildings and paths in the building and outside the building, bridge and

Track, etc. At the same time, in the hierarchical spatial arrangement, objects within the same cluster may also be classified as not belonging together and thus not measured against each other.

The presence of a 3D model can yield great advantages in terms of dimensioning from the aspect of hierarchies. So things can be measured against each other, which have a special meaning. This meaning does not have to be available for the whole room. This results in significant advantages in terms of readability, memory and computing time. Thus, under certain circumstances, a distance between two locations can be specified depending on the importance of different ways.

According to claim 6, the validity of the data model is verified and / or updated by 3D measurements. As reference system for the verification measurement, the reference system of the data model is selected so that the objects in the environment are measured against the data model reference system. For example, therefore, the railway serves as a reference system for a railway line. The measurement for verifying the data model is thus carried out in such a way that all objects are measured against the track specified in the data model.

Claims 7 to 12 describe a data carrier with data stored thereon or signal sequence representing data suitable for transmission via a data network, the data representing a program for execution on a data processing system, which implement the method claims described in claims 1 to 6, if the program is running on a computer, computer network or over the Internet. In order to avoid repetition, reference is essentially made to the previously described statements relating to claims 1 to 6 with regard to these subject matters.

Claims 13 to 18 describe a computer system, which is set up as a client-server system so that it implements the method claims described in claims 1 to 6. Again, to avoid repetition, reference is essentially made to the previously described embodiments of the claims 1 to 6.

The computer system is set up with its client-server system in such a way that at least one database containing the position data and all associated attributes is installed on the server. It also contains libraries of standard object models suitable for the relevant applications. The communication between client and server takes place via a suitable protocol, e.g. over the internet.

Claim 19 describes a computer program with program code for performing all method steps according to at least one of claims 1 to 6, when the computer program is executed on a computer, a computer network or via the Internet. A computer program is to be understood as a machine-executable and executable sequence which is embodied by storage, for example on a data carrier or in a volatile or nonvolatile memory module of a computer, or by signals which are sent via the Internet. The computer program does not need to be in an immediately executable form; it can also in a form prepared for installation on a user's computer.

The invention is explained in more detail below with reference to exemplary embodiments and two figures.

Fig. 1 shows the basic elements of the SD data model according to the invention and how the different measurement data are combined on a data processing system. The spatial arrangement of the three-dimensional space is determined by buildings, paths / roads etc., technical installations such as e.g. Components of the control and safety technology and the terrain, to which also vegetation or similar is counted, by all these elements are related to the highly accurately measured tracks and reference points.

The system has data on different aspects. The geometry of the objects and the environment is described by the spatial and possibly temporal data components. The description of the networking of the data makes it possible to bring the data into specific contexts. For example, it is possible to describe the geometric relationships in different geo-models.

Furthermore, it is e.g. possible to describe other relationships through functional networking, e.g. in the simulation of functionally connected signaling systems.

Time-dependent states or attributes of variable objects are summarized in the figure by the terms state / action.

The history of the data and objects described by the data can be tracked by comparing the plans and the model being measured to reality over time.

The different data of different accuracies are suitably combined in a data processing system. Design data is merged with measurement data from terrestrial surveying, image / film recordings and 3D scans. Using the data processing system, it is now possible to perform further measurements in the digital 3D data model in the data processing system.

When validating the data model, measurements are taken between the real scene and the data model (differential measurement method). The Measurement is based on different degrees of accuracy and refers to points, lines and surfaces in the data model, which are attributed to a particularly high accuracy.

In contrast to the prior art, spatial planning is therefore not defined by measured distances between measured measuring points, but by distances between real objects and basis points / lines / surfaces in the data model. These base points / lines / surfaces do not necessarily have to represent real existing objects, so that abstract measurements are possible both in the model itself and between the real environment and the model.

FIG. 2 shows how object categories are divided into different clusters by way of example on the basis of their safety relevance for a rail transport network. Depending on the safety relevance, different requirements for the accuracy of the data are defined. The highest demands on the accuracy of the position data are placed here on the rails. The position of the rails forms the reference point for the spatial arrangement of the entire three-dimensional scenery. Clustering is performed using weights from different areas according to security relevance for sufficiently high accuracy at low data density. The higher the weight, the higher the accuracy requirements of the 3D data model data. The rails and the objects in close proximity to the rails, such. As the elements of the control and safety technology and the platform, are rated in the figure with the weighting of w = 100. The slope area around the track is assigned a weighting w = 70, so it is treated with a slightly lower accuracy. More distant objects such as buildings receive a weighting w = 50, since the safety relevance for operation on the rail transport route is correspondingly lower. The lowest accuracy is given with a weighting of w = 20 for the forest area. In a first example, a 3D data model is to be created for a railway track network. All objects relevant to the route infrastructure should be displayed in it. In addition to the position data of the tracks, this also includes, for example, the position data of the components of the control and safety technology, level crossings, structural objects that limit the gauge on the railway tracks, such as bridges, tunnels, overhead masts, signs, etc., as well as structural objects characteristic of the conversion fields around the railway line and finally natural objects, such as the vegetation.

The 3D data model should refer to the shape of the earth. No projections are made as in conventional methods, but the model follows the earth curvature or at least one underlying geoid model.

Since different applications are to use the data, it is necessary to prepare the data in such a way that users with simple terminals receive the data in another form adapted to the performance of their terminal, than users who have more powerful computers available. For example, a user with a simple navigation device is an employee with a GSM-R terminal that is to be sent to a specific location for maintenance or landscaping. The amount of data and detail of the position determinations must here be much smaller than e.g. for a planner to schedule a new coverage signal at a breakpoint on a PC.

For an examiner who has to decide whether or not a train is allowed to drive on a given route at a given overshoot, other conditions apply with regard to the required detail of the track-like environment, which he can master with a powerful computer.

The database has a uniform 3D model. This uniform 3D model is constructed on the computer primarily according to collected data. In order to have a database suitable for all conceivable applications, the data is collected in such a way that the greatest possible accuracy can be guaranteed. First of all, all available position data must be collected and merged along the route.

The data are available in different systems and types, with different accuracies. Underneath are extensive 3D data generated by different 3D scanning methods, e.g. by laser or ultrasonic or the like Scanning methods were obtained.

As data sources come u.a. in consideration:

- Measurement data from the terrestrial survey

Elevation data from NASA or helicopter flights 2D and 3D drawings from project planning

Photogrammetric data obtained from environmental images taken with a stereo camera, simple camera images, or movies

- 3D scans of laser or ultrasound scanners in the vicinity

In particular, the 3D scans provide a large amount of data that is difficult to interpret. The data consists of an extensive point cloud, whereby each measuring point must be assigned the appropriate object.

The different data should also be standardized and managed in a common data format.

For the railway operation unnecessary data should be filtered out.

In order to reduce the amount of data, which is usually in the range of a few terabytes, and to obtain the fastest possible access to the data, the data should be compressed.

For this purpose it is exploited that the different accuracies of the different measuring methods and the type of measured objects are known.

In a first step, the measurement data of the terrestrial survey are evaluated and the attributes of the measured objects are recorded, which indicate which objects are involved.

The measured data points are compared with the associated reference points of a geoid model, eg with the help of triangulation networks. Depending on the geo-model, it is thus possible to relate both to the geographic model of the earth and to a gravity model. The model of gravity, for example, makes more sense there than a purely geographic model of the shape of the earth, where gravity following the outflow of water must be considered. The objects are divided into different categories. Decisive for the classification of the objects is their safety-related relevance for railway operations. In the case of objects, for example, a distinction is made between objects in the railway infrastructure, structural objects that limit the clearance gauge on the railway lines, structural objects that are characteristic of the surroundings around the railway line, and natural objects such as vegetation. The railway infrastructure includes z. Trackbed, platforms, components of the control and safety technology, level crossings, etc. Structural objects that the

Limit the gauge space on the pathways are e.g. Bridges, tunnels, trolley masts, signs, etc. These object categories are classified according to their safety relevance into different clusters. Depending on the security relevance, different requirements for the accuracy of the data in the respective clusters are defined.

The highest demands on the accuracy of the position data are placed on the rails. The position of the rails also forms the reference point for the spatial arrangement of the entire three-dimensional scenery. The rails are measured with millimeter precision in this example. Accordingly, the objects in the rail cluster are stored with an accuracy of one millimeter in the SD data model.

The elements of the control and safety technology are to be measured in this example with a lower resolution than the rails. Here, the objects of the associated cluster are to be considered approximately to the centimeter. To a lesser extent, the clusters are to be treated with the structural objects that are close to the clearance gauge. More distant objects or vegetation data can be treated with even lower resolution.

Then a first mixing with other data takes place. For this purpose, the data of a Digital Terrain Model (DTM) are merged with the data already collected. Wherever data of the DGM has an accuracy worse than required in the cluster for the object category, the data from the terrestrial measurement, if any, will be used. Where no other data is available or the data of the DGM better fit the object category, the data of the DGM are used.

DGM comes in a variety of forms (e.g., helicopter flights, earth orbit scans, measured terrestrially by surveyors).

Digital terrain models are constructed from 3D data e.g. created using triangulation networks.

When combining the different 3D data from DTM and terrestrial surveying, the correct connections for the triangulation networks between parts of different accuracies must be found. Higher accuracy corresponds to a denser triangulation network. Therefor First, common nodes must be found. In the intermediate area of the nodes, the data points are interpolated.

The density of the nodes and the triangulation networks associated therewith is selected according to the required accuracy for the object categories. Thereby unnecessary nodes are deleted.

This significantly reduces the amount of data in the 3D data model compared to the original raw data. Since, in the merging of the data, a triangulation network defined according to the invention was created with mesh sizes corresponding to accuracy clusters, the data is available in a uniform data format.

Thus, essentially the data model for the railway line has been created. There is a self-contained 3D model whose data volume has been significantly reduced compared to the original measurement data. It is now possible to access the data of the SD model with different applications.

A planner of a new line can use the data to simulate the route realistically. To do this, he accesses the server with the database of 3D data via the Internet from a computer equipped as a client. His planning data are placed in the 3D model with the appropriate accuracy to the associated clusters where they are to be built later in reality. At his client he can use the high computing power to display the planned objects on his monitor to the highest possible accuracy.

A landscape keeper is to cut back vegetation growth at a certain point in the route network. For this, he can also access the data of the 3D model, where it is displayed at which point the vegetation is to be cut back. However, he has no client computer available, but a smartphone that accesses the server via a GSM-R service. Since this terminal does not provide a large display, the data is transmitted to it in a much lower resolution. For this application, therefore, the tracks or other objects in the vicinity of the route must not be displayed in more detail than the vegetation.

The server recognizes for which application the 3D data is requested and calculates the appropriate data for the representation based on the accuracy requirements specified in the object categories for the different applications. The data are thus transmitted to the landscape care worker with a mesh size of the triangulation network or point density of the objects to be displayed which is adapted to the performance of the terminal. The data is displayed on the display unit so that the dot density of the display is resolved according to the required accuracy. This also reduces the required signal transmission bandwidth to such an extent that even devices with low performance or signal transmission bandwidth can receive and display the data.

In a second example, the amount of data is further reduced to further relieve both the database and the terminals as well as the bandwidth for signal transmission.

As in the first example, the data is collected in a 3D database for the 3D model, classified, assigned to clusters and unified.

When merging the data, moreover, the redundant data points are determined and deleted so that each data point is contained only once in the database. For example, the data points of the tracks are at least partially present both in the data of the terrestrial survey and redundantly in the data from a survey in helicopter overflights. Thus, all redundant track data points of the helicopter overflights can be deleted.

When processing the data, data points that are actually useful for the required accuracy of the assigned object category are also removed from the database, but can be calculated with sufficient accuracy from the surrounding data points by an interpolation method. The prerequisite for this is that this data can be calculated within the required accuracy from the surrounding data points. In the case of linear and area-type structures, methods of linear interpolation, such as, for example, are used for the calculation. As polylines and triangulations. For curved lines, for example, custom curve and area splines are used for interpolation. In order to have information regarding the accuracy of the interpolations available, weightings are assigned to the measured data points in addition to measurement and display tolerances. The respective weights and their combination in a common function characterize the validity of a measure of a particular measurement method in the 3D model. When interpolating points or areas between the data points used for the data model, basis vectors of the underlying local coordinate system are used. These generally refer to points of different accuracy. The position vector to an interpolated point is represented as a linear combination of the basis vectors, which are weighted with the accuracy of the associated base points. This results in a linear combination of the weights, which provides information about the accuracy of the interpolated point or the interpolated surface.

To achieve the required accuracy in the clusters different methods are used.

- triangulations

- Method with simply curved surfaces

- Double-curved surface method.

The weighting of the interpolated points on surfaces takes place e.g. on the mean deviation of measured points to the interpolation surface. In addition, frequently occurring and standardizable objects are determined, and it is checked whether the measured values accurately reflect such an object within the scope of measuring accuracy. If so, these data points are removed from the database. The 3D model is equipped here with the information that a corresponding algorithmic object is to be calculated there.

In the case of a 3D model of a railway track, this is possible, for example, with railway sleepers or trolley masts, signals, level crossings, etc. These objects do not necessarily have to be stored in the database by concrete measuring points. If a data analysis reveals that the measurement values correctly represent such an object within the scope of the measurement accuracy, the object can be simplified in a simplified form by an algorithmic object. be calculated or calculated, which considerably reduces the amount of data in the database. Only if there are deviations from the algorithmic objects that go beyond a measure to be selected are the measurement data points of such objects not replaced by algorithmic objects.

Another example describes how a generalized, hierarchical measuring principle is used in a 3D model according to the invention.

In the hierarchical measuring principle, spatial planning within the model is measured by distances between objects of the same relevance clusters. Only measurement relations are allowed which relate to matching objects, e.g. Platform and track, track and level crossing, buildings and paths in the building and outside the building, bridge and track, etc.

For emergency planning or acute emergencies, the hierarchical measuring principle is used to investigate escape routes to determine whether and how they can be passed on to the rescue teams and their equipment, or their vehicles. In addition to the field data, data on routes within buildings such as train stations, etc., will be included in the database and utilized.

For normal vehicles, for example, paths for passability are examined by specifying the width of the paths or the distances between the path defining buildings or the like. be determined.

For all-terrain vehicles, routes that are off the beaten path for vehicles are also examined, provided that the terrain is sufficiently smooth. For this, distances between objects of other object categories are taken into account, e.g. Trenches, electricity pylons, etc. The appropriate measurement hierarchies for the respective measurement task are then measured along the possible paths and then checked to see if the vehicle dimensions allow them to be used.

In another example, there are still data from 3D scans, which were integrated into the database. These measured data often have a high level of noise. Thus, a further reduction of the amount of data is possible if the measurement data are released from the noise component. The noise component represents a certain proportion of measured data that can not be assigned to actually existing objects. These measurement points are filtered out by comparing the scan data with the other data in the 3D model. Deviations from the other data recorded in the data model can occur. when objects have been added or removed. Furthermore, objects already contained in the data record may have changed, for example, if a curb of a platform edge has shifted. Such data is detected using object and edge detection. Data points that can not be meaningfully assigned to such analysis methods are either examined in greater detail by the processor or are automatically classified as noise and deleted. A closer examination will be carried out especially in the areas of the 3D model where high safety requirements are imposed. For this purpose, the potential noise points are stored and presented to the user with a corresponding note on the display unit.

However, this may not cover all the objects along the way. The timeliness of the data can not be permanently guaranteed because of new buildings or changes in the track. Therefore, in another example, the existing measurement data is compared with reality using 3D measurements. This comparison of data only captures differences between the new measurement data and the existing data points of the previously created 3D model. As soon as new measuring points emerge which can not be reconciled with existing data points in the context of the error tolerances, the data are stored and compared to the

Classified object category and the accuracy cluster. These data points may need to be further explored as they may provide clues to security-related changes.

Furthermore, it is possible to replace the existing data in the data model by the new scan data where the accuracy requirements of the previous data points were not met, if the accuracy of the scan data is better.

The scan data are then freed from their noise.

The examples given are not restricted to the railway sector for application. Rather, they are also transferable to other fields of application. Roads for motor vehicles can also be used as traffic routes, out of town or in the city. Also for describing the infrastructure of other means of transport, e.g. Airfield, port facility, etc., the invention is transferable.

Claims

claims

1 . Method for measuring, evaluating and displaying a three-dimensional environment and the objects located in this environment, wherein the position data and attributes to the environment and the objects are obtained from different data sources, combined in a data processing system to a database, processed and at least one display unit characterized in that the amount of data in the database is reduced by dividing the environment and the surrounding objects into categories of different relevance and the accuracy of the position and time coordinates of the data points for the different object categories depending on their relevance, whereby the spatial density of the measurement data is chosen to be higher for object categories with high relevance than for object categories with lower relevance, and depending on the performance of the display the data is forwarded to the terminal in such a way that the point density of the representation is adapted to the performance of the display unit or of the terminal, the data being displayed on an output unit such that object categories with low relevance are displayed with a lower resolution as object categories with higher relevance.

2. The method of claim 1, wherein the amount of data is reduced by:

- noisy points are detected and deleted from a 3D scan, and / or

- redundant points are determined and deleted, and / or

- Valid points are deleted when they are made redundant by an interpolation method, and / or

frequently occurring objects are replaced by algorithmic objects, and / or

- Points are deleted with improper accuracy, and / or

- High accuracy points are deleted so that the clusters can be scaled up with low security relevance. 3. The method according to at least one of the preceding claims, wherein the data points in addition to measurement and representation tolerances also weights which indicate the validity of a measure of a particular measurement method in the 3D model.

A method according to at least one of the preceding claims, wherein the spatial arrangement of an environment around a distance of a rail network for track-guided traffic is defined by the distance of the objects to the track, being selected as a measure of the relevance of the object data safety aspects.

Method according to at least one of the preceding claims, wherein a hierarchical spatial arrangement is defined within the model, which is determined by distances between objects of defined togetherness, wherein only measurement relations are permitted which relate to matching objects.

Method according to at least one of the preceding claims, wherein the validity of the data model is verified and / or updated by 3D measurements by selecting the reference system of the data model as the reference system for the measurement so that the objects in the environment are measured against the data model reference system become.

Data carrier with data stored thereon or signal sequence representing data suitable for transmission over a data network, the data representing a program for execution on a data processing system for evaluating and displaying a three-dimensional environment and the objects located in that environment, wherein the position data and Attributes to the environment and the objects come from different data sources, the program merges the data in the data processing system to a database, processed and displayed on at least one display unit, characterized in that the program is designed so that it reduces the amount of data in the database by dividing the environment and the surrounding objects into categories of different relevance and the accuracy of the position and time coordinates of the data points for the different object categories depending on their R elevanz, whereby the spatial density of the measured data for object categories with high re- levance is higher than for object categories of lower relevance, and depending on the performance of the display unit, passes the data to the terminal in such a way that it adapts the point density of the display to the performance of the display unit or terminal, where the data is represented on an output device as representing low-relevance object categories with a lower resolution than object categories of higher relevance.

A data medium having data stored thereon or a signal sequence representative of data transmission over a data network, according to claim 7, wherein the program is arranged to reduce the amount of data by

- Detects noisy points from a 3D scan and deletes, and / or

- Detects and deletes redundant points, and / or

- Deletes valid points if they are made redundant by an interpolation method, and / or

Replaces frequently occurring objects with algorithmic objects, and / or

- deletes points with inappropriate accuracy, and / or

- Deletes high accuracy points so that the clusters can be scaled up with low security relevance.

9. Data carrier with data stored on it or for the transmission over a data network suitable, data representing signal sequence according to at least one of the preceding claims 7 to 8, wherein the program is designed so that it assigns the data points in addition to measurement and representation tolerances and weights , which indicate the validity of a measure of a specific measurement method in the 3D model. 10. Data carrier with data stored thereon or for the transmission over a data network suitable, data representing signal sequence according to at least one of the preceding claims 7 to 9, wherein the program is designed so that it is the spatial arrangement of an environment around a distance of a rail network for track-guided traffic Defined by the distance of the objects to the track, where it selects security aspects as a measure of the relevance of the object data.

1 1. Data carrier with data stored on it or data transmission signal sequence suitable for transmission over a data network according to at least one of the preceding claims 7 to 10, wherein the program is designed to define within the model a hierarchical spatial arrangement defined by distances between objects Togetherness is determined, where it only allows measurement relations that relate to matching objects. 12. Data carrier with data stored thereon or data transmission signal sequence suitable for transmission over a data network according to at least one of the preceding claims 7 to 11, the program being designed to verify the validity of the data model by 3D measurements and / or or updated by serving as the frame of reference for the measurement, the reference model of the data model, such that the program measures the objects in the environment against the data model reference system.

13. Computer system, as a client-server system, arranged such that the client evaluates a three-dimensional environment and the objects located in this environment and displays it on its display unit, the position data and attributes relating to the environment and the objects originating from different data sources, wherein the server merges the data in at least one database in the server into a database, processes it and transmits it to the client, characterized in that the server is set up in such a way that it reduces the amount of data in the database by defining the environment and the the objects in the environment are divided into categories of different relevance, and the accuracy of the position and time coordinates of the data points for the different object categories varies depending on their relevance, whereby the spatial density of the measurement data is higher for object categories with high relevance than for object categories with never Of greater relevance, and depending on the performance of the display unit of the client, it passes the data to the client in such a way that it adapts the point density of the representation to the performance of the display unit of the client, presenting the data on the output unit as object categories with ringer relevance with a lower resolution than object categories with higher relevance.

The computer system of claim 13, wherein the server is arranged to reduce the amount of data by

- Detects noisy points from a 3D scan and deletes, and / or

- Detects and deletes redundant points, and / or

Replaces frequently occurring objects with algorithmic objects, and / or

Deletes points with improper accuracy, and / or

15. Computer system according to at least one of the preceding claims 1 3 to 14, wherein the server is set up so that it assigns the data points in addition to measurement and representation tolerances and weightings that indicate the validity of a measure of a particular measurement method in the 3D model ,

16. The computer system according to claim 1, wherein the server is set up to define the spatial arrangement of an area around a track of a track-based rail network by the distance of the objects to the track, using it as a measure of the relevance of the object data security aspects selects.

7. The computer system according to claim 1, wherein the server is set up to define within the model a hierarchical spatial arrangement that is determined by distances between objects of defined togetherness, whereby it only permits measurement relationships, relate to the matching objects.

18. Computer system according to claim 1, wherein the server is set up in such a way that it verifies and / or updates the validity of the data model by 3D measurements by using as reference system For the measurement, the reference model of the data model is used so that the program measures the objects in the environment against the data model reference system. 19. Computer program with program code for carrying out all method steps according to at least one of claims 1 to 6, when the computer program is executed on a computer or a computer network.