DE102004045562A1 - Data-processing method for handling remote-sensing data picked up by an earth satellite offers multiple processing routines - Google Patents

Abstract

Defined requests are entered via an interface (2a) so as to put remote-sensing data to use. In a linkage device (3), a partial portion of processing routines is selected and interlinked by using defined requests, so that selected processing routines (A-F) can complete defined requests by processing the remote-sensing data. Independent claims are also included for the following: (A) a method for selecting remote-sensing data; (B) a method for automatic geo-referencing of remote-sensing data; (C) a system for processing remote-sensing data picked up by earth satellites; and (D) a system for selecting remote-sensing data.

Description

The The invention relates to a method and an arrangement for selection of remote sensing data, in particular remote sensing data, the taken from a satellites.

Remote sensing data form the basis for diverse Investigations of the surrounding environment, starting with the detection of Mineral resources via satellite-based mapping the earth to the collection of environmentally relevant data. Thus, can a large Circle of users use this data useful.

at existing systems for researching remote sensing data, the for are suitable for a particular purpose, is currently a high level of knowledge about satellites and remote sensing data provided by the user. With the growing Diversity of data inevitably results in an increase in selection criteria.

The Networking reception, archiving and data distribution Remote Sensing Data and Data Products Generated From It Will Become National and internationally in the next Years increase. This will increase the amount of data from remote sensing data available be different for solving Issues can be used. It will be beyond that resistant new remote sensing missions initiated so that the data volumes are expected to increase even further. On the other hand it will for the Users of remote sensing data and data products increasingly difficult the overview of the existing possibilities to retain and select the right data and products.

It is an object of the present invention, a method and a Specify arrangement for selecting remote sensing data that it enable, from a variety of available Alternatives to identify those who meet the requirements closest to the user comes. In this case, reproducible results should be achieved and Expert knowledge will not be required.

• • Aufgabenprofildaten verwendet, die Aufgaben und/oder Ziele einer Nutzung von Fernerkundungsdaten definieren,
• • automatisch zumindest eine physikalische Dateneigenschaft, insbesondere eine Auflösung, von Fernerkundungsdaten den Aufgabenprofildaten zugeordnet, sodass zumindest eine Anforderung an die Nutzung von Fernerkundungsdaten definiert ist,
• • Dateneigenschaften von verfügbaren Fernerkundungsdaten unter Verwendung der zumindest einen Anforderung geprüft und festgestellt, ob und/oder bis zu welchem Grad mit den verfügbaren Fernerkundungsdaten die zumindest eine Anforderung erfüllt werden kann, und
• • ein Ergebnis der Feststellung automatisch ausgegeben.

A method is proposed for selecting remote sensing data, in particular remote sensing data acquired from a terrestrial satellite. It will be

• • use task profile data defining tasks and / or goals of using remote sensing data,
• Automatically automatically assigns at least one physical data property, in particular a resolution, of remote sensing data to the task profile data so that at least one request for the use of remote sensing data is defined,
• Data properties of available remote sensing data using the at least one requirement checked and determined whether and / or to what extent with the available remote sensing data, the at least one requirement can be met, and
• • automatically output a result of the determination.

Under a selection of remote sensing data will both the selection already existing remote sensing data as well as the selection of future remote sensing data Understood. Accordingly, "available data" will include available data in the future Understood. With the requirements, in particular a requirement profile, and with the result of the aforementioned finding is the selection possible. Accessing the data (such as retrieving and loading data) can immediately afterwards or later. It can also access be omitted, for. B. if it is determined that the cost of access are unexpectedly high.

at the physical data properties are in particular one or more of the following data properties: local resolution an observation object acquired by remote sensing data, temporal resolution when capturing the object to be observed, spectral resolution of Remote sensing data and / or radiometric resolution (eg, number of pro Picture element available standing digital memory bits).

The task profile data is z. B. generated in a dialogue with a user of the remote sensing data. For example, the user specifies a number of specifications for which tasks the remote sensing data should be used. Such statements are z. For example, the temporal-spatial dynamics of a phenomenon to be detected by evaluation of the remote sensing data (eg a phenological course of vegetation, a temporal change of electromagnetic radiation emanating from an observed surface, or a chlorophyll concentration of waters). Furthermore, the user can specify, for example, the (eg scientific) field of work and / or the economic or administrative area of activity (eg creation of maps for a residential community) in which the remote sensing data is to be used. In this case, the information required by the user can be entered sharply or out of focus. In order to be able to make such statements, a user interface with a Mask, which allows him to enter fuzzy information.

Corresponding the one mentioned above as an example physical data properties can meet the requirements Usability of the remote sensing data in terms of spatial, time, spectral and radiometric resolution, d. H. it will at more complete fulfillment the requirements only uses data with correspondingly limited resolution.

• • mit dem Nutzer interaktiv Unstimmigkeiten abklären kann und/oder
• • auf Wunsch ein korrigiertes Aufgabenprofil erstellen kann. Das korrigierte Aufgabenprofil kann das bisherige Aufgabenprofil ersetzen und das Verfahren kann wie oben beschrieben erneut mit dem korrigierten Aufgabenprofil ausgeführt werden.

In particular, defined rules are used in the definition of the requirements. A requirements profile describing the requirements to be met by the remote sensing data or the remote sensing data processing system (including processing routines) may be created. In addition, the details of the task profile (and optionally a user profile, see below) can be checked for consistency, so that the system:

• • interactively clarify discrepancies with the user and / or
• • can create a corrected task profile on request. The corrected task profile can replace the previous task profile and the method can be executed again with the corrected task profile as described above.

optional can also a user profile are created, eg. B. again in a dialogue with the user: The user enters through a series of details Information Profile. These include e.g. Information on intended use (ie to the thematic environment of the user's work), to his knowledge in terms of remote sensing data usage and processing and / or the decision-making and action potential of an institution that he represents in the context of the remote sensing task. The input of the requested information can the user through allows a defined input mask be over he sharp (precise) or make blurry statements. It is also possible that the user, physically based criteria (e.g., temporal-spatial Resolution) for use indicates the remote sensing data.

If a user profile is created, the at least one physical Data property of remote sensing data the task profile data and the user profile data are assigned, so that the at least defines a requirement for the use of remote sensing data is. If a user profile is not created, the task profile may be be taken as well founded and as a basis for the determination serve the result.

optional the user can preferences in terms of of the evaluation and selection problem. The preferences can be used, in particular, to determine the influence that Constraints (such as cost of using remote sensing data and data processing products, data availability, deployment period, accuracy and quality the data, degree of fulfillment a task profile defined by the user) on the selection result exercise.

The Invention has the advantage that theoretically unlimited amounts from remote sensing data can be automatically selected data that for one specific purpose. No expert knowledge is required because even by inexperienced users entered information due the assignment of physical data properties to physical defined requirements for the use of remote sensing data. Especially The many different satellite sensors have different Type (e.g., with different temporal, spatial, and spectral resolution). Data that has data properties that meet the requirements so can now from an automatic data processing system z. B. be searched in one or more databases. It can too additional Taken into account become (s.o.).

• a) Multitemporale Erkundung, d.h. identische Landschaftsobjekte werden zu unterschiedlichen Zeitpunkten erfasst.
• b) Multistationäre und -sensorielle Messung, d. h. identische Objekte werden bei unterschiedlicher Beobachtungsgeometrie (Bahn-/Flughöhe, Blick-/Beleuchtungswinkel des Satelliten) und durch unterschiedliche Sensoren (z. B. passive/aktive Sensoren) beobachtet.
• c) Multispektrale Erkundung, d. h. Objekte werden in verschiedenen Spektralbereichen gemessen (charakteristische Spektralsignaturen).
• d) Multiple Nutzbarkeit, d. h. Nutzung der Daten zur Lösung verschiedener Aufgaben sowie in Kombination mit Daten anderer Herkunft zur Ableitung komplexer Informationen.

Added to this is the multifunctionality concept of remote sensing, which the user usually can not or not fully consider in a selection of remote sensing data. The multifunctionality concept has the following aspects:

• a) Multitemporal exploration, ie identical landscape objects are acquired at different times.
• b) Multistationary and sensorial measurement, ie identical objects are observed with different observation geometry (orbital / flight altitude, viewing / illumination angle of the satellite) and by different sensors (eg passive / active sensors).
• c) Multispectral exploration, ie objects are measured in different spectral ranges (characteristic spectral signatures).
• d) Multiple usability, ie use of data to solve different tasks and in combination with data from other sources to derive complex information.

Due to the invention, the multi-functionality concept is automatically taken into account. For example, as part of the task profile data, the user inputs information indicating that properties of particular landscape objects should be evaluated with the remote sensing data (see point a). When assigning the physical data properties can therefore be automatic It should be borne in mind that the same landscape object can only be observed by the same satellite with temporal interruptions and recurrence (see point b).

general is formulated in a development of the invention, a physical Model implemented that allows the movement of satellites relative to the surface of the planet (or other object) in the definition of the requirements to take into account. In particular, you can further information, eg. B. over the orientation of a sensor of the satellite, considered in the model become. The for For example, the model required input information stored in a database and / or can be entered by the user become.

• – Informationen über verschiedene Sensoren in oder an demselben Satelliten und
• – physikalische Eigenschaften von Sensoren, mit denen Fernerkundungsdaten aufgenommen werden (z. B. Spektralbereiche, in denen ein Sensor Strahlung detektieren kann, spektrale Empfindlichkeiten der Sensoren).

The same applies to the points c) and d) of the above list. In particular, the definition of the requirements may take into account alternatively or additionally:

• Information about different sensors in or on the same satellite and
• - physical properties of sensors used to acquire remote sensing data (eg spectral ranges in which a sensor can detect radiation, spectral sensitivities of the sensors).

As from point d) can different tasks or even different task profiles lead to the selection of the same remote sensing data and processing routines. Therefore It may happen that already the requirement profile similar or is identical, although the tasks are different. In particular are therefore multiple assignments of the same physical properties possible to different tasks.

• • problemorientierte Auswahl von Fernerkundungsdaten,
• • Anforderungen an die Fernerkundung hinsichtlich der räumlich, zeitlichen, spektralen und radiometrischen Auflösung,
• • Navigation durch Datenbestände und
• • problemorientierte Produktdefinition (d. h. Definition eines Ergebnisses einer Weiterverarbeitung der Fernerkundungsdaten)

In particular, the invention may be used for a consulting system that helps the user to precisely define one or more of the following processes, or performs the process (s) in whole or in part or prepares for execution:

• • problem-oriented selection of remote sensing data,
• • requirements for remote sensing in terms of spatial, temporal, spectral and radiometric resolution,
• • Navigation through databases and
• • problem-oriented product definition (ie definition of a result of a further processing of the remote sensing data)

at A preferred embodiment of the invention are those from the Task profile data generated requirements for the use of remote sensing data used to control data processing of the remote sensing data. In particular, the requirements are made of one or more databases corresponding remote sensing data is requested and loaded and will controlled the data processing of the loaded data. Especially can use the requirements at least one predefined Processing routine for processing the remote sensing data to be selected. Further embodiments of such a control are possible.

Especially can during the exam, Whether the at least one requirement can be met, a requirement profile (with the at least one requirement) and a performance profile of Remote sensing data are compared. The performance profile of the remote sensing data Information about it, which services (already today or in the future) can fulfill available remote sensing data. Especially have the service profile and the requirement profile the same Types and / or the same number of categories of properties the remote sensing data, so for example, a pairwise Comparison of the categories listed in the corresponding categories Properties possible is.

• • ein Arbeitsgebiet, wobei die Fernerkundungsdaten für eine Tätigkeit auf dem Arbeitsgebiet genutzt werden sollen,
• • die geometrische Auflösung der Fernerkundungsdaten,
• • die zeitliche Auflösung der Fernerkundungsdaten,
• • die radiometrische Auflösung der Fernerkundungsdaten,
• • die spektrale Auflösung der Fernerkundungsdaten,
• • optional: einen Ort und/oder ein Gebiet, dessen Fernerkundungsdaten ausgewertet werden sollen,
• • eine oder mehrere Randbedingungen (z. B. Preis der Nutzung von Fernerkundungsdaten).

For example, the requirement profile and the performance profile each have the following categories:

• • an area of work where remote sensing data is to be used for work in the field,
• The geometric resolution of the remote sensing data,
• The temporal resolution of the remote sensing data,
• The radiometric resolution of the remote sensing data,
• The spectral resolution of the remote sensing data,
• • optional: a place and / or area whose remote sensing data is to be evaluated,
• • one or more boundary conditions (eg price of using remote sensing data).

The performance profile may be stored, for example, in one or more databases. Accordingly, the performance profile itself may be construed as a characteristic of a set of remote sensing data or a remote sensing data processing system. A remote sensing data processing system is understood to mean a set of remote sensing data and associated processing routines that can be used to process the remote sensing data into a defined product. The set of performance profiles describes the possibilities, the best to meet the requirements.

at comparing the requirement profile and the performance profile the degree of fulfillment z. B. from the similarity (Distance) between requirements and performance profile. It can at a comparison of differences in the compared Profiles performed become. It can Differences and matches Weighted differently in each category become. The emphasis is on preferences of the user and / or weights defined by the user. In particular, the differences are according to the weighting categorically mapped to a numeric scale, so for each evaluated A numerical value is formed, and becomes the sum of the numerical values are added to an overall result of the comparison. In this way receives one rating.

In the evaluation of the available Remote sensing data or remote sensing data processing systems may be used alongside the degree of fulfillment also additional Conditions (e.g., costs of using remote sensing data).

The belonging to these boundary conditions Values are z. B. Properties of Alternative Remote Sensing Data or reconnaissance systems and are shared with data properties provided. Which boundary conditions can be included in the evaluation is z. B. predefined.

In the rating can both sharp (for example, defined by a numeric value) as well as fuzzy data (eg one by a term like "high resolution" or "low resolution") Resolution "defined requirement) incorporated, especially in the boundary conditions but also in the tasks the task profile data. In particular, at first all fuzzy data be converted into crisp data and then the evaluation done. Alternatively, a so-called fuzzy evaluation method (English: fuzzy logic) be applied. An example will be with reference to the attached figures received.

Especially can be the degree of a match one of the requirements (in the requirement profile) with one of the Properties (in the performance profile) determined in the following manner become: possible Requirements in at least one of the categories will become one math function (element) associated with, by each of the functions potential Properties in the performance profile one rating each assigned. If there is a concrete requirement and becomes the requirement compared with a property in the performance profile, the Degree of agreement by Determines that the valuation function is assigned to the assigned function is determined and as a degree of agreement is used. The requirements are particularly blurred formulated, d. H. arise for a plurality of different properties one each Degree of agreement, which is not zero.

The Automation of the comparison of performance profile and requirement profile allows a comparison of the requirement profile with a variety of performance profiles. Therefore, you can theoretically unlimited many performance profiles on their consistency be compared with the requirement profile. Or it may be the degree their agreement be determined so that it is possible is, the performance profile with the highest degree of agreement with the requirement profile. So z. For example, a ranking all available standing alternatives.

• – eine Schnittstelle zur Eingabe von Aufgabenprofildaten, die Aufgaben und/oder Ziele einer Nutzung von Fernerkundungsdaten definieren,
• – eine Zuordnungseinrichtung, die mit der Schnittstelle verbunden ist und die ausgestaltet ist, automatisch zumindest eine physikalische Dateneigenschaft, insbesondere eine Auflösung, von Fernerkundungsdaten den Aufgabenprofildaten zuzuordnen, sodass zumindest eine Anforderung an die Nutzung von Fernerkundungsdaten definiert ist,
• – eine Prüfeinrichtung, die ausgestaltet ist, Dateneigenschaften von verfügbaren Fernerkundungsdaten unter Verwendung der zumindest einen Anforderung zu prüfen,
• – eine Feststellungseinrichtung, die ausgestaltet ist festzustellen, ob und/oder bis zu welchem Grad mit den verfügbaren Fernerkundungsdaten die zumindest eine Anforderung erfüllt werden kann, und
• – eine Ausgabeeinrichtung, die ausgestaltet ist, ein Ergebnis der Feststellung automatisch auszugeben.

Also proposed is an arrangement for selecting remote sensing data, in particular remote sensing data acquired from an earth satellite, the arrangement comprising:

• An interface for entering task profile data defining tasks and / or goals of using remote sensing data,
• An association device which is connected to the interface and which is designed to automatically associate at least one physical data property, in particular a resolution, of remote sensing data with the task profile data so that at least one request for the use of remote sensing data is defined,
• A test device configured to test data properties of available remote sensing data using the at least one request,
• A determination device configured to determine whether and / or to what extent the at least one requirement can be met with the available remote sensing data, and
• An output device configured to automatically output a result of the determination.

The invention further comprises a computer program for selecting remote sensing data, wherein program code means of the computer program are designed to execute the method according to the invention in at least one of its embodiments. In particular, the program code means may be stored on a computer-readable medium. Furthermore belongs to the Scope of the invention, a data carrier and / or a computer system, on which a data structure is stored, which executes the computer program after loading into a working and / or main memory of a computer or computer network.

embodiments The invention will now be described with reference to the accompanying drawings. there can any individual features or combinations of features of embodiments combined with the previously described embodiments of the invention become. The individual figures of the drawing show:

1 a particularly preferred arrangement for the selection of remote sensing data,

2 a flow chart illustrating a particularly preferred embodiment of the method according to the invention,

3 schematically an assignment of physical data properties to the task profile data,

4 a relationship between geometric resolution of a satellite sensor and the temporal resolution (frequency) in the recording of observation objects,

4a a legend in 3 and 4 used numbers and abbreviations explained,

5 a schema that represents a hierarchical breakdown of criteria

6 an arrangement for the processing of remote sensing data, in particular the in 1 contains shown arrangement, and

7 a set of functions that can be used to check a fuzzy requirement for agreement with a sharply defined property.

In the 1 The arrangement shown has an interface IN for inputting data and for outputting results of the arrangement. Via the interface IN, a user of the arrangement can enter information (for example, inter alia via a keyboard) and can be informed about the result (for example, via a screen and / or a printer). The information entered by the user is used in the arrangement, in particular when creating a task profile.

The Interface IN is connected to an allocation device MA, their task is the assignment of physical data properties to the information entered by the user. The allocation device MA is directly connected to a memory device DL and also respectively indirectly via a testing device MD, a weighting device MG and an evaluation device MB connected to the memory device DL. The storage device DL can be optional as well be connected to an archive AF.

In the memory device DL are in particular data properties remote sensing data and processing routines for processing the remote sensing data stored. The remote sensing data itself are stored (if available) in the AF archive. alternative may also be another memory structure with, for example, only one memory device be provided in both the data properties, the data itself and the processing routines are stored.

With reference to 1 and up 2 a particularly preferred embodiment of the method according to the invention is described.

Based The user specifies a number of details via the interface IN, for which Tasks the remote sensing data should be used. such Indications are e.g. the temporal spatial Dynamics of a phenomenon to be detected. For this For example, the interface IN may provide the user with a mask for entering the Provide information that allows him to enter fuzzy information allows.

In addition, can the user over enter the IN interface information about himself those in the interface IN and / or in the allocation device MA a user profile can be created. additional Information for the User profile be obtained in other ways, for example by access on already stored information about the user, for example are stored in the memory device DL. It can be for one Users (eg interactive in dialogue with the user) an individual User profile created and saved, the z. B. Information to the preferred field of application of remote sensing data, knowledge from earlier Inquiries, geographical and / or temporal resolution and boundary conditions. By using the saved user profile may be for the future a pre-selection of criteria and data attributes are made.

However, the user profile can also be dispensed with, for example if it is an experienced user. In this case will the user profile z. B. disabled by the user or the arrangement. In this case, the task profile is taken as a profile created by an experienced user and this is taken into account in the selection of remote sensing data.

Of the Users can already classify the information entered by them in predetermined criteria. For example, he can with the Information about the tasks a temporal and / or spatial resolution specify the remote sensing data to be used.

Farther the user may have preferences about the Enter interface IN, through which he weights for the rating while of the selection process. In particular, so can the weight constraints (such as cost of use the remote sensing data, data availability, period of deployment remote sensing data, accuracy and / or quality of data, Minimum level of fulfillment the tasks defined by the user with the selection result).

In the allocation device MA can the Task profile and the user profile are combined and are assigned the physical data properties so that a requirement profile is generated. The requirement profile describes that from the remote sensing data processing system (or remote sensing data) to be fulfilled Conditions.

Preferred dimensions are the information of the user for the task profile and for the user profile checked for consistency, so that the arrangement with the user interactive clarify discrepancies and / or on request can create a corrected task profile.

The allocation means MA preferably uses the inference method in the combination. Fuzzy or sharp values are processed using predefined rules, as described with reference to 7 will be explained with reference to an example. The user gives z. For example, enter information such that words are defined as values of linguistic variables. The linguistic variables can then z. B. are reproduced by blurred elements in the profiles.

in the Case of creating the rules by an expert, the Check rules using examples (for example, training scenarios) and possibly. iterative be adjusted. For example, it can be a neural network with used a plurality of neuronally-linked fuzzy logic nodes be trained and not a priori unambiguously provides specific result when using defined input data become. On the other hand, if the user enters sharply worded requirements, which do not allow alternatives, the result is from the outset firmly.

Now is from the test facility MD at least one stored in the memory device DL performance profile compared with the requirement profile of the allocation device MA. Preferably, a plurality of performance profiles with the requirement profile compared. The set of performance profiles describes the possibilities which exist to meet requirements. As a result, the test equipment MD over for example the allocation device MA and the interface IN (or via a other interface) to the user to deliver that request be fully or partially fulfilled with the performance profile. In particular, results a degree of fulfillment from the similarity (Distance) between requirements and performance profile. It is also It is conceivable that an optimized solution will be returned from performance profiles (See below the description of the valuation institution MB) The weighting device MG leads, for example (as already described) a pairwise comparison of properties through, filed in corresponding categories of profiles are.

The Evaluation Facility MB examines and evaluates the performance profile or the performance profiles taking into account the requirement profile. In the evaluation of existing remote sensing data processing systems go next to the degree of fulfillment preferably also the boundary conditions. For this purpose, the boundary conditions considered as components of the profiles, so z. Legs or multiple categories in the profiles reflect the constraints. It may be stipulated which boundary conditions in the evaluation flow in can. On the basis of the evaluation result, a (according to the degree of agreement ordered) Ranking of all available performance profiles or a ranking of the remote sensing data defined by the performance profile or remote sensing data.

By Hierarchical staggering of the criteria will increase the complexity of the request minimized and achieved a weighting of the criteria.

By a hierarchical staggering of selection criteria, the complexity of the input for the user can be reduced and a weighting of the criteria can be achieved. The array may rely on classifications of data properties that allow mapping of the properties to the selection criteria. 5 shows an example of the hierarchical structure. The total destination of data usage is shown with block OZ. The overall goal may include one or more sub-goals Z1, Z2, Z3, etc. For each of the sub-goals Z1, Z2, Z3, etc., criteria K1, K2, K3, etc. are defined, for each of which sub-criteria U1, U2, U3, etc. are defined. This allows the user to specify individual subcriteria, so that the overall goal can be derived from it. The results of the interrogation of the subcriteria can then be summarized and an overall result determined that allows to describe the overall objective. In doing so, user-defined weights of the criteria can be taken into account.

By a mapping of user requirements to a matrix (in lines and column ordered matrix elements serving as the target reach matrix can be designated) in a preferred embodiment, those entered by the user Information defined requirements processed mathematically become. By evaluating this matrix using the MADM method (Multiple Attributes Decision Making Method) is next to the plausibility check a Objectification of the request allows. The MADM method is z. In Zimmermann, H.-J .; Gutsche, L. (1991): "Multi-Criteria Analysis - Introduction to the theory of decisions in multiple objectives ", Springer-Verlag, Berlin, Heidelberg, New York, London, Paris, Tokyo, Hong Kong, Barcelona, Budapest. ISBN 3-540-54483-6 (Berlin), ISBN 0-387-54483-6 (New York). On the parts this publication, which describe the MADM method is hereby incorporated by reference taken.

With a target achievement matrix (in particular as described above has been checked for consistency of the information or criteria entered by the user is) can then by the arrangement of a request to the storage device DL triggered whether there is a corresponding performance profile.

The presented solution allows easier access to remote sensing data and provides a essential simplification of the search. This also results an improved utilization of existing data.

According to the MADM method the criteria defined in the requirements profile or Properties according to their (user-specified) weighting compared with each other and with regard to the target. Especially an eigenvector of the goal attainment matrix is formed and becomes using the eigenvector the consistency of the user verified criteria. optional can the criteria are corrected. The goal may be in the form of a vector whose elements meet the criteria so that by comparing the eigenvector with the target vector, a distance vector can be determined. For example, by forming the amount the distance vector becomes the distance and thus a measure of the degree determined, with which the goal with the criteria can be reached.

The Application of the MADM method for setting up and evaluating a target achievement matrix facilitates the decision making process and allows the objectification of the Requirements by applying expert knowledge. The expert knowledge but does not have to be present at the user, but is already flowed in before. The method results in a higher match of found data realized with the expectations of the users. The described embodiment thus represents an interface to the data management system, which the basis of the MADM method the respective requirement profile of the User as possible compares well with the data descriptive properties.

From the arrangement can (eg according to the evaluation result ordered) the most appropriate remote sensing data processing systems or remote sensing data for the determined requirement profile and the given boundary conditions be issued. additionally the user can over possible Deviations between task profile and requirement profile or selected performance profile be informed and / or about appropriate alternatives are informed.

has the array on the archive AF, in the remote sensing data stored are the most appropriate to the requirements profile corresponding remote sensing data can be automatically output and / or can the most appropriate processing routines be called automatically. This makes it possible to automatically create a processing job generate and process the remote sensing data.

3 shows an example of the assignment of physical data properties (here: the geometric resolution) to the task profile data (here: work areas or work situations for which the remote sensing data should be used). The geometrical resolution is assigned in the figure subject areas or work areas in the wide middle column, which is overwritten with "dimension" and which is divided into several columns.The geometric resolution of the remote sensing data is indicated in the figure by the right, with "pixel resolution" overwritten column. The numbers in this column are footnotes that are each assigned to a satellite and / or a sensor of a satellite. This sensor or satellite provides remote sensing data with geometric resolution. Thus, for example, Fer For the work area "Environmental Systems in the field of Climatology" with the topic "Terrain Climate" needed, one finds the remote sensing data with suitable geometric resolution by determining the footnotes horizontally at the same height as the term "Terrain Climate" in the right column "Pixel Resolution" , here eg 24 (satellite IKONOS with sensor "MS", ie multispectral) A suitable geometric resolution (namely a better geometric resolution) is also provided by all satellites or sensors whose footnote is in the right column below the found footnote.

So, if the user enters the field of "Environmental Systems in the field of Climatology" with the topic "Terrain Climate", a linear geometric resolution of 4m and / or an areal geometric resolution of 10 m ^{2 is} assigned to this information and included in the requirement profile. These resolutions can be seen in the "Mapping" column in 3 read off. In the figure, different map scales are logarithmically plotted in the left column, which correspond to the geometric resolutions.

For the selection suitable remote sensing data is in addition to the geometric resolution of Remote sensing data, in particular the temporal resolution of Data of importance. It decides on the question of whether the extraction real-time data made realistically close to the observed process can be. In addition, the physical potential of the sensor to fulfillment be questioned and / or simulated the task.

In the 4 the relationship between geometric resolution of sensors and the temporal provision of data is shown. For the sake of understanding, it is pointed out that a satellite only reaches the same relative position to a specific point on the earth's surface after a certain period of time (period duration). Therefore, the same observation area can be observed from the same point of view only after the period has been repeated. However, it is possible that part of the observation area may be re-observed earlier by the satellite, but from a different point of view and / or as part of another observation area.

In 4 the period of different satellites in days is plotted on the horizontal coordinate axis. The period is defined here as temporal resolution. Again, the numbers entered in the diagram should be understood as footnotes indicating satellites or sensors of satellites. However, the indication of the period duration does not preclude the fact that the selection of the remote sensing data or remote sensing data processing systems also takes into account the fact that a part of an observation area and the like may be considered. U. earlier than at the end of the period can be observed again by the same sensor or satellite.

One recognizes 4 in that the geometric resolution achieved by the satellites or sensors is approximately functionally related to the temporal resolution. The functional relationship is represented by the dotted line, which decreases with increasing period to smaller values of geometric resolution. Ie. As the period increases, the geometric resolution becomes finer. Preferably, this relationship is also used in the selection of remote sensing data. In particular, the consistency of the information input by the user can be checked on the basis of the relationship as described above. It can be revealed contradictions in this information, ie it can, for. B. an unreasonable fine required by the user geometric resolution can be discovered.

According to the above described relationship increases with an increase of the spatial resolution (pixel resolution) usually the width of the scannable image strip on the earth's surface (swath) so that less area coverage is reached. Nevertheless, there are comprehensive data to be able to record must be a higher number of circulations of the satellite. This inverse behavior of geometric and temporal resolution is represented by the dotted line. The problem with the inverse behavior regarding a satellite may be able to circumvented by the fact that data comparable satellites (z. Satellites with a sensor providing similar data) be used. In the near future, even at the same time started several identical satellites. Such constellations of satellites and their recorded remote sensing data can then in the method according to the invention considered become.

Again, in 4 the columns "Mapping" and "Scale" are entered.

In the 6 shown arrangement for processing remote sensing data has z. B. the in 1 represented interface IN and the in 1 shown storage device DL. With RS is designated a device that another, based on 1 already described selection of remote sensing data or remote sensing data processing systems. For example, the device RS may be the allocation direction MA, the test device MD, the evaluation device MB and the weighting device MG have. Further, a controller CR is provided for controlling the processing of remote sensing data and for controlling the selection of the remote sensing data or systems. The control device CR is connected to the device RS and to a device PRO for processing the remote sensing data. The arrangement PRO is in turn connected to an interface or storage device RD, from which the device PRO receives remote sensing raw data for further processing.

When operating the entire in 6 shown arrangement is first, for example, as based on 1 and 2 described a selection of a remote sensing data processing system. Information about the selection is transmitted from the device RS to the control device CR. For example, the selected power profile may also be transmitted from the storage device DL to the device PRO. The controller CR controls the operation of the device PRO such that the remote sensing data is processed in accordance with the selected remote sensing data processing system by the device PRO. The arrangement PRO can request corresponding raw data from the interface or storage device RD. In particular, the arrangement PRO comprises a chain of processing routines for processing the raw data, wherein the already pre-existing processing routines are combined in accordance with the selected remote sensing data processing system.

As already described in the evaluation of available Remote Sensing Data or Remote Sensing Data Processing Systems also fuzzy properties of data and / or fuzzy boundary conditions (such as the price of using remote sensing data) become. In one embodiment The evaluation can be a so-called fuzzy evaluation method with a "blurred Logic "(English: Fuzzy Logic) be applied.

at Fuzzy logic becomes linguistic variables as input variables used. These are understood as variables whose values are not numbers (as with deterministic variables) or statistical distributions (as with random variables), but linguistic or linguistic Constructions (also called terms). These terms are blurred Elements assigned to a set of elements or by these elements represents. The elements can also referred to as fuzzy value functions of a base variable become.

One embodiment will now be described on the basis of the constraint "price". It's on other base variables, namely Boundary conditions and / or physical properties of the remote sensing data transferable. Is it a physical property (such as the geometric resolution) can even go through an input of a sharp value (eg 30 m) the user a range of values (20-40 m) or a fuzzy statement (eg again a linguistic Value like "high Resolution ") as a requirement be assigned. It is thus generated a blur, which has the advantage has that now deviates from the sharp value properties (eg 40 m) in the performance profile lead to the result that meets the requirement or at least to a certain extent.

In the exemplary embodiment, the basic variable "price" has five fuzzy elements which mathematically describe the degree of affiliation μ of a linguistic price statement (see, for example, FIG. 7 : "Very cheap", "cheap", "acceptable", "expensive" and "too expensive".) The elements are each functions that describe the degree of affiliation of the respective linguistic statement to values on a price scale.

The elements always have the value 1 in a value range of the price scale. From this value 1, the elements steadily decrease (eg, linearly) to zero. The boundary elements (in 7 : "Very cheap" and "too expensive") only steady on one side to zero, while the middle elements on both sides steadily decrease to zero. The value range with the value 1 and the steady (ie in particular not abrupt) decrease express the blurring of the affiliation of a price statement to a price. None of the elements overlap another element, so the range of values of both elements in the overlap area is 1. However, this may be the case in other embodiments of the method.

By specifying the linguistic statement "very favorable" by the user, the corresponding element is selected as a property of the price for the requirement profile If, when comparing the property, a certain price is compared in the performance profile, the element can be assigned the degree of membership for that price is determined and interpreted as weight If the degree of affiliation is 1 (in the example of 7 for the price range from 0 to x1), full compliance exists. In the area of the gradual decrease or increase in the degree of affiliation, the degree of affiliation lies between 0 and 1 (in the example of 7 for the price range from x1 to x2). Therefore, for a price in this range, the match is rated with the membership grade between 0 and 1. Where the price belongs 0, there is no match, ie the request is not fulfilled.

The elements may be predetermined and / or properties of the elements may be entered by the user. The elements do not have to (as in 7 shown) have constant value ranges of the degree of affiliation.

Much more Any other designs of the elements are possible, the however, preferably each value of the base variable uniquely membership degree (the weight).

Claims (13)

Method for selecting remote sensing data, in particular remote sensing data obtained from a terrestrial satellite were taken, where • Task profile data used, the tasks and / or goals of a use of Define remote sensing data, • automatically at least one physical data property, in particular a resolution of Remote sensing data is mapped to the task profile data so that at least one requirement for the use of remote sensing data is defined • Data properties from available Remote sensing data using the at least one request checked be determined and whether and / or to what degree with the available remote sensing data the at least one requirement can be fulfilled, and • a result the determination is issued automatically. The method of claim 1, wherein in addition to the task profile data Also user profile data can be used, the properties of a Describe user, and wherein the at least one physical Data property of task profile data and user profile data is assigned, so that at least one request for use is defined by remote sensing data. Method according to one of the preceding claims, wherein except the data properties of available Remote sensing data also available Processing routines for the further processing of the remote sensing data be checked using the at least one request and it is determined whether and / or to what extent with the available remote sensing data and the available Processing routines are met at least one request and a result of the determination is automatically output. Method according to one of the preceding claims, wherein a physical model of the motion of satellites relative to surface describes an object observed in the remote sensing, where the physical model used in the definition of the requirements and wherein the remote sensing data from at least one of the satellites be recorded or were. Method according to one of the preceding claims, wherein the usage requirements generated from the task profile data of remote sensing data to control a data processing of Remote sensing data can be used. A method according to the preceding claim, wherein requested remote sensing data corresponding to the requirements and being loaded and being the data processing of the loaded data is controlled according to the requirements. Method according to one of the preceding claims, wherein during the exam, Whether the at least one requirement can be met, a requirement profile and compared a performance profile of the remote sensing data with the same performance profile and requirement profile Species and / or an equal number of categories of properties the remote sensing data and wherein in the corresponding one another Categories dropped properties compared in pairs become. A method according to the preceding claim, wherein Differences and matches in the individual categories are weighted to varying degrees and where dependent weighted by a user's preferences. Method according to one of the two preceding claims, wherein potential Requirements in at least one of the categories mathematical functions be assigned, by each of the functions possible Properties in the performance profile one rating each is assigned and where a degree of a match one of the requirements with one of the properties is determined by that from the assigned function the rating number is determined and as a degree the match is used. Computer program for the selection of remote sensing data with program code means configured, the method according to the claims 1 to 9. Computer program with program code means according to the preceding claim, stored on a computer-readable medium are. disk or computer system on which a data structure is stored, after loading into a working and / or main memory of a Computers or computer network the computer program according to claim 10 or 11 executes. Arrangement for the selection of remote sensing data, in particular from remote sensing data taken by a geosatellite were, the arrangement comprising - an interface (IN) for entering task profile data, the tasks and / or Define goals of using remote sensing data, - An allocation device (MA), which is connected to the interface (IN) and which is designed, automatically at least one physical data property, in particular a resolution, from remote sensing data to the task profile data, so that at least one requirement for the use of remote sensing data is defined - one test equipment (MD), which is designed to provide data characteristics of available remote sensing data using the at least one requirement to consider - a determination device (MG, MB), which is designed to determine if and / or up to what degree with the available Remote sensing data that meets at least one requirement can, and - one Output device (IN) designed to be a result of Issue automatically.