WO2009013271A1

WO2009013271A1 - Method to make automatically available cartographic data in a container crane system, container crane system and control program

Info

Publication number: WO2009013271A1
Application number: PCT/EP2008/059534
Authority: WO
Inventors: Hans-Peter Kagerer
Original assignee: Siemens Aktiengesellschaft
Priority date: 2007-07-26
Filing date: 2008-07-21
Publication date: 2009-01-29
Also published as: DE502008000357D1; EP2043940A1; ATE457017T1; EP2043940B1; DE102007035034A1

Abstract

At least one satellite image of a container crane system is created with a resolution which can be predetermined to automatically make available cartographic data in a container crane system. The order of containers in the features describing the container crane system is extracted from the satellite image by means of a pattern recognition method. A container crane system-specific topographical map is generated from the extracted features. Said map is made available for at least one control unit of the container crane system, which unit is assigned to at least one transport device.

Description

description

Method for the automatic provision of cartographic data in a container crane system, container crane system and control program

Newer container crane systems have a very high degree of automation. However, due to environmental conditions and dimensions, automation solutions encounter limits in certain areas of a container crane system. For example, human intervention is still required due to long distances, weather influences, material movements and deformations. In addition, measuring systems must be constantly maintained and calibrated at regular intervals.

"Rubber-tired gantries" or "straddle-carriers" are controlled by humans and are very flexible in terms of field of application. Rubber-tired gantries, in particular, require very precise driving in an area where containers are stacked, and mistakes made by a driver during steering can have dangerous consequences or at least result in significant damage to property.

Systems for the prevention of driving errors are based on a precise knowledge, especially of container stacks. Control units for transport facilities of a container crane system are therefore supplied with cartographic data so that lanes can be reliably detected. For structural or operational changes within a container crane system, the cartographic data for the control units must hitherto largely be tracked manually. Control systems for container cranes usually work with fixed reference points within the container crane system. For this purpose, a system coordinate system is defined. Position details of individual container cranes then refer to corresponding coordinates. Basically, floor plans of a container crane system are available at the operator. However, these plans often deviate from the real world or are not precise enough to accommodate positioning accuracy requirements in the range of a few centimeters.

The present invention has for its object to provide a method for automatically providing cartographic data in a container crane system and to provide a suitable implementation of the method and an adapted container crane system.

This object is achieved by a method having the features specified in claim 1, a container crane systems with the features specified in claim 6 and a

Control program solved with the features specified in claim 7. Advantageous developments are specified in the dependent claims.

An essential aspect of the method according to the invention is that for the automatic provision of cartographic data in a container crane system at least one satellite image of the container crane system is created with a predefinable resolution. By means of a pattern recognition method, features describing a container arrangement in the container crane system are extracted from the satellite image. From the extracted features, a container crane-specific topographical map is generated. The card is used at least for one control unit of the Contai- Nerkrananlage made available, which is associated with at least one transport device. With fundamentally up-to-date cartographic data at any time, it is much easier to react to changes on a container crane site than with previous solutions.

Advantageously, the satellite image is provided by an internet-based service. In this way, a particularly economical access to required for generating cartographic data starting material results.

According to an advantageous development of the present invention, the extracted features are checked for plausibility by comparison with reference values. This serves to provide additional quality assurance of the automatically acquired cartographic data. For example, the reference values are determined during at least one crane crossing over at least one container stack. Alternatively or additionally, the reference values can be queried in an inventory management system in which container inventories are recorded.

The method according to the invention is preferably implemented by a control program for the automatic provision of cartographic data in a container crane system. The control program has at least one code section which can be loaded into a main memory of a computer, and in the execution of which the above-described steps are executed or initiated when the control program runs in the computer.

The present invention will be explained in more detail with reference to the drawing. It shows FIG. 1 shows an application environment for acquiring cartographic data according to the prior art;

FIG. 2 shows a flow chart for a method according to the invention for providing cartographic data in a container crane system,

FIG. 3 shows a satellite recording of a container crane system provided by an internet-based service.

An absolute position determination of transport devices and containers within a container crane system has hitherto been carried out by complex manual measurement. In this case, based on a GPS position determination differential GSP position determination method (DGPS) are used. The transport devices used are supplied with surveying results determined in this way. Due to a high manual processing proportion, an increased susceptibility to errors results.

DGPS uses a fixed reference station that receives signals sent by GPS satellites. Since the geographical position of the reference station is known with very high accuracy, an error of a GPS position determination can be calculated. For this purpose, a respective error contribution of a GPS satellite whose signals can be received, individually determined and transmitted by the reference station by radio to DGPS receiver within a given region.

A DGPS receiver decodes received GPS satellite signals and, in addition, reference station correction signals. With In the latter case, the DGPS receiver corrects errors in the GPS satellite signals, thereby increasing the accuracy of the position detection. In many cases, a receiving antenna for the correction signals is already integrated in a GPS antenna. If the radio disconnects from the DGPS transmitter, the receiver switches to a normal GPS mode without correction, but then loses the accuracy advantage. Depending on the quality of the receiver and the correction data, the achievable accuracy is between 0.3 and 2.5 m in relation to length and width data and between 0.6 and 5 m in relation to height information. High-quality systems additionally evaluate a carrier phase and achieve accuracies of only a few mm.

The correction signals of a reference station are usually transmitted by means of radio in the VHF range or by means of a mobile radio connection to the DGPS receiver. Since the erroneous determined transit times of the GPS satellite signals change only slowly, a transmission of the correction signals via VHF or mobile radio connection is non-time critical.

In addition to stationary reference stations, geostationary satellites can emit correction signals which are generated by local reference stations, for example continent-specific. Examples include the systems EGNOS in Europe, WAAS in America and MSAS in Japan.

Alternatively or in addition to DGPS-based methods, position data can be determined by "learning trips" with a transport device, for example, a crane can be moved manually in its operating area, whereby the position data along the travel route are recorded and stored. Figure 1 shows an exemplary arrangement of a container stack 102 within a container crane system. To start a learning process, a crane operator positions a crane 101 on a starting point 111 at the beginning of a stacking area. The crane operator stores a first reference position via an input device, for example push-button or control panel. As a result, the crane driver drives with the crane 101 along the container stacks and thereby detects several intermediate positions 112-114. This is especially important if the container stacks are not arranged along a straight line.

Lastly, an end position 115 of the container stacks is stored. Thus, a setpoint route for the crane 101 along the container stack considered here is detected. In addition, other cranes used along the container stack considered here must also be provided with information on the recorded target route, otherwise separate "learning trips" would have to be carried out for the other cranes. In addition, position corrections can be made by position sensors installed in transport vehicles or by absolute encoders provided in drive units.

In accordance with the method illustrated in FIG. 2, for the automatic provision of cartographic data in a container crane system, first of all a satellite image of the container crane system is created with a predefinable resolution (step 201). Features describing a container arrangement in the container crane facility are then extracted from the satellite image by means of a pattern recognition process (step 202). From the extracted features, a container crane-specific topographical map is generated (Step 203). The card is finally made available for a control unit of the container crane system, which is assigned to a transport device, for example a crane.

The method is preferably implemented by means of a control program which has at least one code section which can be loaded into a main memory of a computer and in the execution of which the above-described steps are carried out when the control program runs in the computer. In this way, a container crane system with a unit for retrieving at least one satellite image of the container crane system can be realized with a predefinable resolution. In addition, the container crane system has a unit for the extraction of a container arrangement in the container crane descriptive characteristics from the satellite image and a unit for generating a container crane-specific topographic map. In addition, a unit for providing the card is present at least for a control unit of the container crane system.

For pattern recognition, the acquired satellite image is segmented by generating content-related regions as well as aggregating neighboring pixels in accordance with a specific homogeneity criterion.

For complete segmentation, each pixel is assigned to at least one segment. For a coverless segmentation, however, each pixel is assigned at most one segment. In a complete and coverage-free segmentation, therefore, each pixel is assigned to exactly one segment. If each segment forms a coherent area, there is a coherent segmentation. Automatic segmentation can be pixel, edge or region oriented. In addition, in the case of model-based methods, a specific form of objects to be recognized is assumed, or, in the case of texture-based methods, an internal homogeneous structure for objects to be recognized is additionally or alternatively taken into account. The approaches to automatic segmentation mentioned here can also be combined with each other.

In the case of pixel-oriented segmentation, the decision is made for each individual pixel as to whether it is to be assigned to a selected segment. This decision may be influenced by the environment of the respective pixel.

Edge-oriented segmentation searches for edges or object transitions in the image to be segmented. Depending on the algorithm used, discrete edge sequences may be combined to form closed edge features that include objects.

Pixels belonging to an edge can be determined by means of the Sobel operator, the Laplace operator or a gradient search. The determined pixels are subsequently completed by means of an edge tracking algorithm. An example of an edge tracking algorithm is the Live Wire method developed by E. Mortensen, WA Barrett and JK Udupa. This method determines an optimal path from a start point to a destination point. Optimal in this context means that the path between start and end point always leads over the strongest edge pixels. This problem can be solved, for example, with a breadth-first search known in computer science. In a region-oriented segmentation, sets of points are considered as a whole in order to find contiguous objects in this way. For region-oriented segmentation, the region Growing, Region Splitting, Pyramid Linking or Split and Merge can be used.

Model-based segmentation is based on a model of objects to be recognized; in the present case, for example, containers or cranes. For model-based segmentation, the Hough transformation is preferably used. Using the Hough transformation, points are combined into lines or circles by mapping them in a parameter space.

In the case of texture-oriented segmentation, use is made of the fact that, although some image objects do not have a uniform color, they have a uniform texture. For example, an object may have grooves that appear in a photograph as alternating streaks of dark and light color.

The satellite image acquired for automatically providing cartographic data in a container crane system is preferably provided by an internet-based service. For example, a service such as Google Earth could be used in principle provided the up-to-dateness of the provided image material is guaranteed.

The features extracted from the satellite image can additionally be checked for plausibility by comparison with reference values. For example, the reference values can be determined during a crane crossing over a container stack or stored in an inventory management system. With the method described can be a complex, manual capture and processing of cartographic data accounts in a container crane system. On the one hand, this prevents input errors in existing data acquisition systems. On the other hand, a considerable saving of time is realized, since it is only necessary to travel long distances in the context of "learning trips" in exceptional cases.

By eliminating "learning journeys" both accident hazards and downtimes due to incorrect or outdated cartographic data are minimized.Additionally, automation systems in a container crane system can be kept up to date using the described procedure.Firstly, sensory equipment can also be functionally kept to a minimum to safeguard security-related functions, such as collision protection.

FIG. 3 shows a satellite image of a container crane system. Dotted lines 301-303 exemplify a selected stack area that can be recognized by the method described above.

The present invention is not limited to the application example described here.

Claims

claims

1. A method for automatically providing cartographic data in a container crane system, in which - at least one satellite image of the container crane system is created with a predefinable resolution,

A feature of the container system describing features in the container crane system is extracted from the satellite image by means of a pattern recognition method, - a container crane-specific topographical map is generated from the extracted features,

- The card is made available at least for a control unit of the container crane system, which is associated with at least one transport device.

2. The method of claim 1, wherein the satellite image is provided by an Internet-based service or within a data network.

3. The method according to any one of claims 1 or 2, wherein the extracted features are checked by comparison with reference values for plausibility.

4. The method of claim 3, wherein the reference values are determined at least one crane crossing over at least one container stack.

5. The method of claim 3 or 4, wherein the reference values are stored in an inventory management system.

6. Container crane with

A unit for retrieving at least one satellite image of the container crane system with a predefinable resolution,

- A unit for the extraction of a container assembly in the container crane descriptive features of the

Satellite image by means of a pattern recognition method,

A unit for producing a container crane-specific topographical map from the extracted features, a unit for providing the map, at least for a control unit of the container crane system, which is assigned to at least one transport device.

7. Control program for the automatic provision of cartographic data in a container crane system with at least one section of code that can be loaded into a working memory of a computer, and during its execution

A creation of at least one satellite image of the container crane system is initiated with a predefinable resolution,

To extract features describing a container arrangement in the container crane system from the satellite image by means of a pattern recognition method,

A container crane-specific topographical map is generated from the extracted features,

- The card is made available at least for a control unit of the container crane system, which is associated with at least one transport device when the control program runs in the computer.