Method and apparatus for monitoring industrial facilities

Abstract

In a method for monitoring an industrial facility (2) including facility objects, the following steps are carried out: The method includes the steps of: receiving (S1) a video stream (VS) of a facility (2) generated with the aid of a video camera device (3), the video stream (VS) comprising a temporal sequence of two-dimensional image data; assigning (S2) two-dimensional image components (PT) of the video stream (VS) to facility objects (A', B', C', D') with the aid of a three-dimensional (3D) model (CAD) of the facility (2), the facility objects (A', B', C', D') being assigned respective object characteristics (XYZ); determining (S3) recording characteristics of the image data depending on the image components (PT) assigned to one or more facility objects (C') and the respective object characteristics (XYZ), the recording characteristics comprising a position, an orientation and / or a zoom ratio of the video camera device (3) supplying the video stream; and outputting (S4) the video stream (VS') together with the determined object characteristics (XYZ) and / or recording characteristics.

Description

[Technical field]

[0001] The present invention relates to a method and a device for observing or monitoring an industrial installation.Furthermore, a computer program product is proposed for enabling a programmable device to perform the method for monitoring an industrial installation. [Background technology]

[0002] Generally, industrial facilities such as factories, especially chemical industries, distribution systems for raw materials and processed products, and logistics industrial systems are visually monitored or recorded using video cameras. To ensure monitoring of industrial facility objects that interact with each other in a complex manner, it is customary to use a large number of surveillance cameras monitoring their respective spatial areas and providing video streaming data.

[0003] It is desirable to be able to reliably evaluate such real-time video data and to be able to intervene if an anomaly is detected in the behavior of the facility. Conventionally, video camera devices are distributed over the spatial area to be monitored and are, for example, rotatably mounted and equipped with a zoom function. The generated video streams often contain no information beyond the video data, and the specific location of the surveillance camera providing the video stream and the settings of this camera remain unknown. Conventionally, therefore, monitoring of the displayed video or film is performed, in particular, manually. Summary of the Invention [Problem to be solved by the invention]

[0004] In particular, it is desirable to simplify the visual surveillance of industrial installations using video streams, and in particular to provide more than just image data or video sequences, since typically in already installed surveillance systems of industrial installations, only image data is available.

[0005] The assignment of 2D image data components of an industrial facility to an existing three-dimensional (3D) facility model is described in N. Navab et al:Cyclicon: a software platform for the creation and update of virtual factories", 1999 IEEE, pages 459 - 463, 1999 7th IEEE International Conference on Emerging Technologies and Factory Automation. Proceedings ETFA '99, ISBN: 0-7803-5670-5. Cyclicon allows the representation of a 2D projection of the 3D facility model superimposed on the 2D image data. However, this requires a pre-assignment of a known configuration of surveillance cameras to a specific 3D model data and vice versa. Furthermore, the user may be required to mark specific pre-defined image points in the 2D image data to be able to perform the superimposition of the video streams.

[0006] In the prior art, a prerequisite for image analysis was that the origin and contextual content of the 2D image data must be predefined. In practice, this is not common, especially since there are many video streams from different cameras whose recording settings, orientations, and other properties are not documented. Furthermore, cameras are often not fixed, but rather rotatable. This makes it virtually impossible for users to manually annotate video stream data while the video stream is running.

[0007] SUMMARY OF THE PRESENT EMBODIMENTS Against this background, it is an object of the present invention to provide an improved method for monitoring an industrial facility, and in particular a large number of video camera systems and facility objects. [Means for solving the problem]

[0008] Accordingly, a method for monitoring an industrial facility including a facility object is proposed, the method comprising the steps of: receiving a video stream of the facility generated with the aid of a video camera device, the video stream including a temporal sequence of two-dimensional image data; assigning the two-dimensional image components of the video stream to facility objects with the aid of a three-dimensional model of the facility, whereby respective object characteristics are assigned to the facility objects; determining recording characteristics of the image data in response to the image components assigned to one or more facility objects and respective object characteristics, the recording characteristics including a position, an orientation, and a zoom ratio of a video camera device capturing the video stream; Outputting the video stream along with the determined object characteristics and / or recording characteristics. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0009] The reception of the video stream takes place in particular at a monitoring location remote from the location of the video camera device, in that the video camera device is mounted at an unknown location during the process of receiving the video stream, at the monitoring location the recording characteristics are not directly observable but are advantageously derived from two-dimensional image components and a three-dimensional model of the facility.

[0010] The method steps, in particular the allocation and determination steps, are preferably performed in an automated manner, which may occur in a computer-implemented manner in the background, without the user explicitly invoking the recording and object properties.

[0011] An industrial facility is understood to mean in particular a system of technical installations which interact with each other and function, for example, for the production, processing, relaying of substances. Technical installations can be understood as facility objects or facility parts. By way of example, infrastructure facilities for water and / or wastewater, re-cooling facilities, power distribution facilities, fuel depots and supply facilities are / are suitable as industrial facilities. In one embodiment, the industrial facility to be monitored is a facility related to the chemical industry.

[0012] In this method, from image data obtained from a video camera device such as a digital surveillance camera, or from a video stream comprising a temporal sequence of two-dimensional image data, respective contextual information related to facility objects captured in the camera and / or video stream is ascertained. Outputting the video stream and / or the two-dimensional image together with the determined object or recorded characteristics allows for an improved classification of the condition of the monitored industrial facility.

[0013] Recording properties are understood to mean in particular properties which can be assigned to a video stream or to the individual 2D image data and which depend on the creation of the respective image. These are for example image angle, excerpt, recording position or orientation of the video camera device, etc. Exposure settings are also considered as recording properties. Camera properties may also be a term used.

[0014] It is also possible to determine the recording characteristics taking into consideration environmental conditions such as brightness and weather.

[0015] An image component may be, for example, an excerpt from a full image or an image dataset created by filtering.

[0016] For example, a three-dimensional facility model, such as a CAD model, can be used, where the facility objects, i.e. components of the industrial facility, are provided with known object properties. Suitable object properties include, among others, the position, size or dimensions, alignment or orientation, and a list of sub-components or material and surface properties of the respective facility objects. Furthermore, possible object properties include, but are not limited to, a list of assignable static or dynamic data, such as, for example, material and design details, fluids and substances, flow rates, temperatures, pressures, or parameters such as the operation of fittings and controlled variables.

[0017] The position, orientation or zoom ratio (zoom factor) of the respective video camera device can be inferred by evaluation of the two-dimensional image components, in particular implemented on a computer. That is, with the help of the proposed method, the recording conditions or video camera devices directed at the image components or specific facility objects can be more specifically characterized as recording characteristics from the two-dimensional image data sequences obtained in the manner of video streams using an appropriate three-dimensional model of the industrial facility. The recording characteristics regarding the orientation, i.e. the direction in which the surveillance camera is facing, and the image angles that can be captured or the currently set zoom ratio of the surveillance camera can be obtained without collecting additional data. The orientation can be specified in a local or global coordinate system.

[0018] In an embodiment, one or more video camera devices have a video capture area that includes the spatial surroundings of the respective video camera device, which is defined by camera properties such as capture angle, panning possibilities, zoom possibilities, resolution, light sensitivity, etc. In the case of a fixedly mounted video camera device with a fixed focus and fixed zoom, the video stream obtained by the video camera device shows a fixed image excerpt and in this respect forms the surrounding video capture area.

[0019] In some embodiments, the industrial facility is located on a spatial site or facility site. Typically, each video capture area is smaller than the factory site. As an example, only multiple overlapping video capture areas of multiple video camera devices cover the factory site. In some embodiments, the video capture areas are on the facility site, but can be located independently of each other or can be spatially separated by non-surveillance areas.

[0020] In one embodiment, the video camera device is positioned within the facility's grounds, and the three-dimensional model of the facility specifically includes the entire facility grounds. The step of determining the recording characteristics then preferably includes the steps of: determining the location of the video camera device within the facility grounds, and preferentially within the video capture area.

[0021] In an embodiment, the position of the video camera device in a global and / or local coordinate system is output.

[0022] The global coordinate system may be relative to the facility's grounds, and the local coordinate systems may be relative to the respective video capture regions. In the environment, the three-dimensional model of the facility covers the facility's grounds in the global coordinate system, and the positions of the video camera devices are known in the global coordinate system, and preferably also in the local coordinate system. In this respect, the video capture regions preferably cover the local coordinate system.

[0023] In this way, the observation and monitoring of the facility is improved. In case of an event, for example an accident, the accident site can be located and reached by personnel more quickly and easily. The assessment of the current facility status or current situation is facilitated by automatic completion of recorded data and / or object data that cannot be obtained from the video stream alone. This is useful not only for the management of accidents, but also in general for the evaluation, optimization and promotion of operator control during the operation of the facility.

[0024] In some embodiments, the surveillance cameras themselves are captured as part of the facility, i.e., as facility objects within the 3D model. In any case, the position or location of each camera can be known and can be used to determine the orientation and zoom ratio.

[0025] Advantageously, the video stream is output together with the object or recording characteristics so that comprehensive image information can be identified, for example, on a mobile terminal equipped with a display device or on a control console for monitoring an industrial facility.

[0026] In this regard, in one embodiment, the video stream is presented along with the object characteristics and recording characteristics with the aid of a display device.

[0027] Monitoring of an establishment is understood in particular to mean the observation, evaluation and optimization through improved information provision. In this respect, the capture or derivation of information about the "monitored" establishment with the help of video stream data is considered as monitoring.

[0028] In an embodiment of the method, the following steps are performed: creating a database for assigning further functional properties and / or context data to facility objects. Functional properties include, for example, process data, temperature or flow materials. In an embodiment, context data can include, inter alia, time data, weather data and / or visibility conditions. Any of the above-mentioned assignable static or dynamic data is suitable as context data.

[0029] The database can in particular be linked to a three-dimensional model of the facility, and with the help of the three-dimensional model data and functional properties, each recorded facility object can be identified from the image components. In particular, based on the identification of the facility object in the image components and the properties of the object stored in the three-dimensional model, such as size, position or orientation, it is possible to determine the position of the camera, its orientation and / or the zoom ratio.

[0030] In particular, image processing methods as mentioned at the outset are suitable here.In particular in certain embodiments of the method, the functional properties and / or the context data are updated in real time.

[0031] In the method, dynamic recording characteristics can be determined, for example, using the temporal sequence of two-dimensional image components of the video stream. It is conceivable to capture the camera movement and zoom speed, which are output as recording characteristics.

[0032] In certain embodiments of the method, in a particularly automated manner, the orientation or zoom ratio of each video camera device supplying the video stream data is altered in response to dynamically or statically captured recording characteristics.

[0033] In an embodiment of the method, the video stream does not contain information data relating to the recording characteristics. In this respect, the recording characteristics are preferably ascertained only by comparison of the assigned two-dimensional image components with the data of the three-dimensional model. In an embodiment, the three-dimensional model data relating to the industrial facility may include the location of the video camera device or devices.

[0034] In one embodiment, the zoom ratio of the video camera device is determined in response to a plurality of time-sequential two-dimensional image data of the video camera device.

[0035] By comparing a plurality of time-sequential image data with the characteristics of an object present in the three-dimensional model, the zoom ratio or zoom speed can be ascertained.

[0036] In one embodiment, the method comprises at least one of the following steps: transmitting the video stream to a processing platform; transmitting the 3D model to a processing platform; and Transmitting the determined recording characteristics from the processing platform to a display device.

[0037] Preferably, the processing platform is designed as a cloud service and is configurable to perform the steps of allocating the two-dimensional image components and determining the respective recording characteristics in a computer-implemented manner in a secure processing environment. The processing platform can be implemented, for example, in the Microsoft Azure environment or using web services from other providers, such as AWS. The respective data is preferably transmitted via a secure encrypted communication protocol.

[0038] In one embodiment, the processing platform is designed as an app running on a terminal, for example a smartphone or a tablet computer, and is configured to perform the steps of allocating the two-dimensional image components and determining the respective recording characteristics in a computer-implemented manner in a secure processing environment of the terminal.

[0039] In one embodiment, the method includes receiving multiple video streams from different video camera devices.

[0040] The steps of allocating the two-dimensional image components, determining the respective recording characteristics and outputting the video streams are preferably performed in each case for the respective video streams and the associated video camera devices using a respective 3D model of the respective industrial facility or facility part, in which in some embodiments the same 3D model can be used, in other embodiments different 3D models are used.

[0041] Video camera devices capture, in particular, different areas of an industrial facility, and the capture direction of each two-dimensional image component of each video stream is defined depending on the orientation, the position and the zoom setting of each video camera. In this respect, it is possible to add additional context to each video stream, such as recording properties and object properties. A representation in which the images of the video streams and the additional properties are superimposed is also conceivable. The additional properties are understood in particular to be camera data and object data.

[0042] An embodiment optionally includes forwarding one of the multiple video streams to a processing platform designed as a cloud service.

[0043] Furthermore, a device is proposed for implementing the above-mentioned and below-mentioned method for monitoring an industrial installation, the device comprising, in particular: Video camera equipment for generating video streams of industrial facilities; a storage device for providing a three-dimensional model of the facility, the storage device having respective object properties assigned to facility objects; A processing device for assigning two-dimensional image components of the video stream to facility objects with the aid of a 3D model and for determining recording characteristics; and / or A display device which assigns and outputs the video stream together with the determined object characteristics and / or recording characteristics to the 3D model, in particular as a superposition. In this case, the video camera device, the storage device, the processing device and the display device are preferably communicatively coupled to each other, and the functionality of the processing device can be provided by a software service in a cloud environment.

[0044] The respective method steps or functionalized devices, e.g. processor units or processing devices, can be implemented from a hardware point of view and / or from a software point of view. In case of implementation from a hardware point of view, the respective functions can be embodied as an apparatus, e.g. a computer or a microprocessor, or as part of an apparatus. In case of implementation from a software point of view, the respective units can be embodied as a computer program product, as an app, as a function, as a routine, as part of a program code or as an executable object, in particular as a software service in a cloud environment.

[0045] Furthermore, a computer program product is proposed comprising computer readable instructions which, when executed by a computer, cause a computer to carry out the above-mentioned method. The computer program product in particular comprises machine readable instructions which, when processed by one or more processor devices of a processing environment, cause the execution of one or all method steps of the proposed method.

[0046] A computer program product, e.g. a computer program means, may be provided or supplied as a storage medium, e.g. a memory card, a USB stick, a CD-ROM, a DVD, or in the form of a file downloadable from a server in a network. This may be done by transmission of the computer program product or a corresponding file containing the computer program means, e.g. in a wireless communication network.

[0047] The embodiments and features described for the proposed apparatus also apply mutatis mutandis to the proposed method.

[0048] Further possible embodiments of the invention also include non-express combinations of the features or embodiments described above or below with reference to the examples, in which case a person skilled in the art may add individual aspects as improvements or supplements to the respective basic form of the invention.

[0049] Further advantageous configurations and aspects of the invention are the subject of the dependent claims and the following described examples of the invention.The invention will now be explained in more detail on the basis of preferred embodiments with reference to the attached drawings, in which: FIG. [Brief description of the drawings]

[0050] [Figure 1] FIG. 1 is a flow diagram illustrating the method steps of one variation of a method for monitoring an industrial facility. [Diagram 2] FIG. 2 is an expanded flow diagram illustrating the method steps of a further variation of the method for monitoring an industrial facility. [Diagram 3] FIG. 3 is a schematic diagram of one embodiment of an apparatus for monitoring an industrial facility. [Figure 4] FIG. 4 shows diagrammatically the allocation of image components to video streams and facility objects to illustrate a method for monitoring an industrial facility. [Diagram 5] FIG. 5 illustrates one embodiment of a display device.

[0051] In the drawings, identical or functionally identical elements are indicated by the same reference numbers unless otherwise noted.

[0052] 1 and 2 are flow diagrams illustrating method steps of an example of a method for monitoring an industrial installation, the method steps being particularly carried out in an embodiment of an apparatus for monitoring an industrial installation as shown diagrammatically in FIG.

[0053] In this respect, Fig. 3 shows a surveillance system or device 1 of an industrial installation 2, which may for example be a complex pipeline system, a factory, a refinery, a production line or a logistics facility. The installation 2, which geographically occupies a large spatial area, can be captured by one or several surveillance cameras 3, 4 designed as digital video cameras. In this case, in particular video stream data is generated in real time. In this case, the video cameras 3, 4 used can be arranged at different positions in the area of ​​the installation 2, for example rotatably or stationarily. It is also conceivable that the different cameras are equipped with an optical zoom. This results in a video stream showing different image excerpts reproducing parts or objects of the installation 2.

[0054] The video stream data is represented by the right-facing arrow in Fig. 3 and is communicated to a local computing device 5, e.g. a control console. The video data provided by the installed camera system does not include information such as the orientation, zoom ratio, focus settings, etc. of the respective cameras 3, 4.

[0055] In Fig. 4, by way of example, a frame of a video stream VS is shown diagrammatically on the left. In this case, an industrial installation 2 is shown simplified with four installation objects A, B, C, D. The video stream VS further comprises a representation of the surroundings of the industrial installation. One of the video cameras 3, 4 can then be focused on an image component PT, for example a pipe system C between two installation parts B, D. To enable a more detailed and therefore more meaningful display for the workers, in a first method step (see Fig. 1), the video stream data VS is received (step S1).

[0056] Method steps S2 and S3 then comprise establishing a relationship between an available three-dimensional (3D) model of the monitored industrial facility 2 or its facility objects or components and the image components of the video stream VS. For this purpose, in step S2, as shown in Fig. 4, each 2D image component PT is assigned to a facility object of the industrial facility using a 3D model, for example a CAD (Computer Aided Design) model of the industrial facility 2.

[0057] FIG. 4 clarifies the assignment by means of dashed arrows. On the right side of FIG. 4, a 3D model of the facility CAD is illustrated. In this case, the facility objects A', B', C', D' present in the 3D model CAD of the facility are assigned object properties. As an example, detailed CAD data are present as object properties. In this case, the object properties can include size, position, orientation in space, and further properties. In the example shown in FIG. 4, the image component PT includes, for example, a pipe C running horizontally and having a certain flowing material. In this regard, object data defining, for example, the permissible surface temperature of the pipe C' can be retrieved from a further database related to the functioning of the industrial facility.

[0058] In step S3, the camera characteristics of the video cameras 3, 4 which provided the image components PT in the video stream VS are ascertained as image recording characteristics. For this purpose, the two-dimensional image data PT reproducing the facility object C, i.e. the pipe system, are compared with the object characteristics present in the three-dimensional facility model CAD. From there, it is possible to ascertain, for example, the position and / or orientation or viewing direction of the cameras providing the image excerpts PT. Furthermore, from knowledge of the positions and the image excerpts PT, it is possible to derive the zoom ratio of the respective camera. In this respect, recording characteristics such as in particular the position orientation and zoom ratio of the video camera arrangement can be captured from the two-dimensional video stream data VS.

[0059] In step S4, these contextual data, including facility object data and recording characteristics, are output together with the video stream.

[0060] In this respect, the proposed method provides for tracking based on image data of the surveillance cameras 3, 4 used. In particular, it is possible to see in real time the camera movements and the zoom settings of the respective cameras. The computer-implemented evaluation of the video stream data and the image processing can be carried out according to known methods, for example as mentioned in connection with the Cyclicon software mentioned at the beginning. However, other algorithms for the recognition of facility objects present in a three-dimensional model based on two-dimensional image data are also conceivable. Also known are algorithms for the recognition of superposition points of a three-dimensional model from a two-dimensional (2D) image. Such algorithms are also available.

[0061] The additional method steps or processes shown in Fig. 2 are carried out in relation to Fig. 3, which shows a cloud-based surveillance system of an industrial facility 2. A preparation step S10 comprises the generation of one or more video streams with the help of video surveillance cameras 3, 4. The video stream data VS, which are shown diagrammatically on the left side of Fig. 4, are supplied to a control console computer 6 via a local computing system. The local computing system 5 receives, for example, the video stream data VS of the different surveillance cameras 3, 4 and prepares them for display on a display. The corresponding data processing steps are carried out locally, which are indicated by braces LKL in Fig. 4 (Fig. 3).

[0062] Since the acquired video stream data VS are present without further context, in particular without knowledge of the position of the used surveillance cameras 3, 4, the video stream data VS or the particular video stream data VS including additional information can be displayed by a suitable terminal for the user USR, for example a tablet computer or a mobile display device 8. For this purpose a cloud platform 7, for example using Microsoft Azure, with corresponding software services 10, 11 is configured. The cloud computing environment CLD is indicated in Fig. 4 (Fig. 3) by a dashed line. In particular, as software services, a 3D facility model service 10 and an application service 11 for communicatively coupling with a user terminal 8 are provided.

[0063] Fig. 3 further shows a database device 9 coupled to the cloud environment CLD via an interface 12. The database 9 contains 3D model data corresponding to the monitored industrial facility 2. These data may be CAD data. Furthermore, the database 9 may contain additional information or context data on facility objects installed in the facility 2. This may for example be the above mentioned information on flowing materials of a particular pipeline. The corresponding 3D facility model or 3D model data is generated or provided beforehand in step S11. For example, such model data is generated during the design or planning of the industrial facility.

[0064] In step S12, the software service 10 instantiates a corresponding 3D facility model based on the 3D model data, e.g. CAD data, retrieved from the database 9. The software service designated at 11 in the cloud environment 7 then essentially executes the method steps S1, S2, S3 and S4 grouped together as processing process S100 of Figures 1 and 2. The dashed double-headed arrow indicates the communication coupling between the terminal 8, i.e. the display device 8 of the user USR, and the software service or app 11.

[0065] Fig. 5 shows a possible embodiment of a display state of a terminal 8. The terminal is for example a display device 8 including a display 14 on which video stream data VS are represented. Furthermore, a user terminal 8 implemented as a display device comprises an operating element 13. By means of the operating element 13, a user USR can call up context information regarding the image data VS' represented on the display 14, for example by means of a touch screen function or other tactile elements.

[0066] As an example, a frame such as that represented on the left side of the video stream VS in Fig. 4 is reproduced in Fig. 5A. Meanwhile, in the cloud environment CLD by the software service 11, the respective facility object data for the facility objects A, B, C, D recognized in the video stream VS are retrieved by the software service 10. Furthermore, these object data are assigned as described with respect to the method step S2.

[0067] For example, the software service 11 may recognize that the pipeline C in the image component represented by the dashed line has a temperature rise that does not comply with the target temperature. When the represented facility area C is selected by the user, an enlarged representation of the area is performed with additional information XYZ specified at 15 added or superimposed, for example as shown in FIG. 5B. The additional information 15 may for example represent warnings, indicating safety events of the monitored facility to the user. Switching back and forth between different cameras can be done by the user based on the camera information ascertained in method step S3. Ascertaining the recording characteristics with respect to orientation and zoom settings simplifies navigation and selection of the video stream images to be displayed using the display device 8.

[0068] Overall, the proposed surveillance method allows enriching pure video stream data with contextual data such as object characteristics stored in a 3D model and tracking the camera's pan and zoom possibilities. The proposed functionality and the augmented representation of the video image data can be provided via a cloud environment and cloud services, and via an appropriate user interface on a display device. Although the invention has been described in more particular detail on the basis of an embodiment, various modifications may be possible. The image components shown as examples, i.e. pipes, may be any facility components. Furthermore, the monitoring may be influenced by further sensor data, e.g. recordings from a spectroscopic camera, which may result in further improvements of the monitoring performance. Instead of CAD data as mentioned for a 3D model, other formats such as OBJ are also possible.

[0069] Reference symbol: 1. Facility monitoring system 2. Factory 3. Video cameras / surveillance cameras 4. Video cameras / surveillance cameras 5. Local Computing Devices 6 Control Console 7. Cloud Environment 8 Display Devices 9. Database 10 3D models 11. Application Services 12 Interface 13 Operational Elements 14 displays 15 Additional Information A, B, C, D Representation of facility objects in the video stream A', B', C', D' Facility objects with object characteristics according to 3D model PT Image Component S1: Receiving video stream S2: A process for allocating 2D image components to facility objects according to the 3D model S3: Process for determining recording characteristics S4 Process to output video stream data with camera / object characteristics S5 Adding context data to the video stream data. S10 Video stream generation process S11 The process of creating a 3D facility model S12 The process of instantiating 3D facility models VS Video Stream XYZ context, camera, or additional information

Claims

1. A method for monitoring an industrial facility (2) including a facility object, the method comprising the steps of: receiving (S1) a video stream (VS) of a facility (2) generated with the aid of a video camera device (3), the video stream (VS) comprising a temporal sequence of two-dimensional image data; A step (S2) of assigning the two-dimensional image components (PT) of the video stream (VS) to facility objects (A', B', C', D') with the aid of a three-dimensional (3D) model (CAD) of the facility (2), where each object characteristic (XYZ) is assigned to a facility object (A', B', C', D'); A step (S3) of determining recording characteristics of the image data according to the image components (PT) assigned to one or more facility objects (C') and the respective object characteristics (XYZ), the recording characteristics including the position, orientation and zoom ratio of a video camera device (3) capturing the video stream; Outputting (S4) a video stream (VS') together with the determined object characteristics (XYZ) and recording characteristics.

2. 2. The method according to claim 1, characterized in that the industrial facility (2) is located within a facility premises and the position of the video camera device (3) within the facility premises is known.

3. 3. The method of claim 2, wherein the video camera device (3) has a video capture area smaller than the facility footprint.

4. 3. The method according to claim 1 or 2, wherein during the step of receiving the video stream (VS), the video camera device (3) is mounted in an unknown position.

5. The method of claim 1 or 2, wherein the object characteristics (XYZ) include the location, size / dimensions, alignment / orientation, component list, material properties and / or surface properties of facility objects (A', B', C', D').

6. Furthermore, the method includes a step (S11) of creating a database (9) for assigning further functional properties and / or contextual data to the facility objects (A', B', C', D'), The method according to claim 1 or 2, wherein the functional characteristics include in particular process data, temperature, flow materials, and the context data include in particular time data, weather data and / or visibility conditions.

7. The method according to claim 6, wherein the functional properties and / or the contextual data are updated, in particular in real time.

8. 3. A method according to claim 1 or 2, characterized in that the allocating step (S2) is further performed in dependence on the functional properties of the facility objects (A', B', C', D').

9. 3. The method according to claim 1 or 2, wherein the three-dimensional model is a computer-aided design model of the industrial facility (2).

10. 3. The method according to claim 1 or 2, wherein dynamic recording characteristics, in particular the camera movement or zoom speed, are determined using the temporal sequence of two-dimensional image components of the video stream (VS).

11. The method of claim 10, wherein the camera movement and / or zoom speed is known in real time.

12. 3. The method according to claim 1 or 2, wherein the video stream (VS) does not contain information data concerning recording characteristics, and the recording characteristics are known exclusively by comparison of the assigned two-dimensional image components (PT) with data of a 3D model (CAD).

13. The method according to claim 1 or 2, wherein the 3D model data comprises the position of a video camera device (3).

14. 3. The method according to claim 1 or 2, wherein the zoom ratio is determined in response to a plurality of time-successive two-dimensional image data of a video camera device (3).

15. Displaying (S5) the video stream (VS) together with the object characteristics (XYZ) and / or the recording characteristics using a display device (8), 3. The method of claim 1 or 2, further comprising:

16. creating (S12) a three-dimensional (3D) model of the facility, in which object properties (XYZ) are assigned to facility objects; 3. The method of claim 1 or 2, further comprising:

17. Further steps of: transmitting the video stream (VS) to a processing platform (7); Sending the 3D model (CAD) to a processing platform (7); and transmitting the determined recording characteristics from the processing platform (7) to a display device (8); and the processing platform (7) is designed like a cloud service (CLD) and configured to perform the step (S2) of allocating the two-dimensional image components and the step (S3) of determining the respective recording characteristics in a computer-implemented manner in a secure processing environment (CLD).

18. The following steps: receiving a plurality of video streams (VS) from different video camera devices (3, 4); 3. The method according to claim 1 or 2, further comprising the steps of allocating (S2) two-dimensional image components (PT), determining (S3) respective recording characteristics and outputting (S4) a video stream (VS) are performed in each case for the respective video stream and associated video camera device (3, 4) using a respective 3D model (CAD).

19. Optionally, the method further comprises the step of forwarding one of the plurality of video streams to a processing platform (7); 20. The method according to claim 18, wherein the processing platform (7) is configured to perform the steps of allocating (S2) the two-dimensional image components and determining the respective recording characteristics (S3) in a computer-implemented manner in a cloud-service-like (CLD) and secure processing environment.

20. A computer program product comprising machine-readable instructions which, when processed by one or more processor units of a processing environment (CLD), cause the execution of one or all of the method steps of the method according to claim 1 or 2.

21. An apparatus (1) for carrying out the method according to claim 1 or 2, comprising: a video camera device (3, 4) for generating a video stream (VS) of an industrial facility (2); a storage device (9) for providing a three-dimensional (3D) model of the facility (2) with facility objects (A', B', C', D') assigned respective object properties (XYZ); a processing device (10, 11) for assigning two-dimensional image components (PT) of the video stream (VS) to facility objects with the help of a 3D model (10) and for determining recording characteristics, and a display device (8) for outputting the video stream (VS) together with the determined object characteristics and / or recording characteristics; wherein a video camera device (3, 4), a storage device (9), a processing device (10, 11) and a display device (8) are communicatively coupled to each other.

Images