A satellite image automated production system based on a space software robot

Abstract

An embodiment of the present invention discloses a satellite image automated production system based on a space software robot. The system includes: The robot center is encapsulated and developed according to the operations required in the satellite image production process; The robot design module selects appropriate functional software robots from the robot center for combination and setting according to the satellite production operation process; The vector spatialization module performs decompression of the original satellite image, extraction of the original image from the image, creation of metadata, mathematical modeling for vector extraction of the image range, and coordinate transformation and projection transformation processing; The space data engine automatically retrieves the production process matching the task initiated by the operator from the robot design module, binds the task parameters input by the operator to the production process to create a task instance, and calls each functional software robot according to the designed production process to process the original image data to produce satellite images. The embodiment of the present invention can improve the production efficiency of satellite images.

Description

Technical Field

[0001] The present invention relates to the technical field of remote sensing big data production, and particularly to a satellite image automatic production system based on a spatial software robot. Background Art

[0002] Satellite image data is the basis for many related works such as urban planning, environmental protection, land survey, and road construction. A series of processing procedures are required from image data acquisition to business application, including image rectification, image registration, data fusion, image mosaicking, block cutting, data quality inspection, and many other steps. Traditional processing methods are basically manual operations, with low efficiency and it is difficult to maintain consistent quality of the results, which to a certain extent affects and restricts the wide application of satellite images. With the rapid development of big data and artificial intelligence technologies, it is necessary to introduce software robots in satellite image production to achieve automatic production. Summary of the Invention

[0003] The technical problem solved by the present invention is: overcoming the deficiencies of the prior art, and providing a satellite image automatic production system based on a spatial software robot.

[0004] The technical solution of the present invention is:

[0005] In a first aspect, an embodiment of the present invention provides a satellite image automatic production system based on a spatial software robot, the system includes: a robot center, a robot design module, a vector spatialization module, and a spatial data engine, wherein,

[0006] The robot center is configured to be encapsulated and developed according to the operations required in the satellite image production process;

[0007] The robot design module is configured to select appropriate functional software robots from the robot center according to the satellite production operation process for combination and setting;

[0008] The vector spatialization module is configured to perform decompression of the original satellite image, extraction of the original image, creation of metadata, mathematical modeling for vector extraction of the image range, and coordinate transformation and projection transformation processing;

[0009] The spatial data engine is configured to automatically retrieve a production process matching the task from the robot design module according to the production task initiated by the operator, bind the task parameters input by the operator to the production process to create a task instance, and call each functional software robot according to the designed production process to process the original image data to produce satellite images.

[0010] Optionally, the robot center includes: a preprocessing software robot, a correction and fusion software robot, an inspection and analysis software robot, a light and color equalization software robot, a mosaicking and finishing software robot, a data cropping software robot, a metadata creation software robot, and a data slicing software robot. Among them,

[0011] The preprocessing software robot is configured to decompress the original image and extract the original metadata to form a list of images to be processed;

[0012] The correction and fusion software robot is configured to correct and perform band fusion on a single original image;

[0013] The inspection and analysis software robot is configured to inspect and analyze the corrected single-image;

[0014] The light and color equalization software robot is configured to adjust the exposure and color of a single image;

[0015] The mosaicking and finishing software robot is configured to splice multiple images and perform transitional finishing processing on the splicing edges;

[0016] The data cropping software robot is configured to crop the mosaicked image according to standard map sheets;

[0017] The metadata creation robot is configured to generate new metadata from the metadata extracted from the original image according to standard map sheets;

[0018] The data slicing robot is configured to process the result data into slice data for publishable image services.

[0019] Optionally, the robot design module includes: a start / end node sub-module, a robot node sub-module, a control node sub-module, an interaction node sub-module, and a production process management sub-module. Among them,

[0020] The start / end node sub-module is configured to control the start and end of the production process;

[0021] The robot node sub-module is configured to select various series of functional software robots in the robot center;

[0022] The control node sub-module is configured to control the production process;

[0023] The interaction node sub-module is configured to pop up prompt messages for operators to confirm for production steps that require manual intervention, and rebind some task parameters during the execution of the production process;

[0024] The production process management sub-module is configured to perform create, modify, and delete operations on the production process.

[0025] Optionally, the spatial data engine includes: a parameter binding sub-module, a task scheduling sub-module, and an execution engine, where

[0026] The parameter binding sub-module is configured to automatically retrieve a production process file matching the task from the robot design module according to the production task initiated by the operator, and bind the task parameters input by the operator to the production process to create a task instance;

[0027] The task scheduling sub-module is configured to detect the total number of task instances in the current system, determine whether there are idle resources to execute the task instance, and if there are idle resources, automatically split the task into multiple sub-tasks that can be executed in parallel, and package the original image data to be produced and the task instance into a task package and send it to the execution engine configured for each hardware resource;

[0028] The execution engine is configured to receive and load the task package sent by the task scheduling module, parse the production process information and task parameters in the task package, start executing from the first node of the production process, and call the corresponding software robot to process the satellite image according to the task parameters of the node.

[0029] Optionally, the parameter binding sub-module is implemented based on the MEF framework, and is specifically configured to use the Container in the MEF framework as the main body for loading the production process, map the parameters that the production task needs to transfer to the Container through the Import interface of the MEF framework, perform parameter assignment and attribute setting on each node of the production process through the Export interface of MEF, and save the selected nodes, assigned parameters, and attribute values of the designed production process as a task instance.

[0030] Optionally, the task scheduling module uses the ActionServlet component in the Struts architecture as the controller, retrieves the production process file matching the assigned task from the robot design module according to the struts-config.xml configuration file that describes the corresponding relationship between the model, view, and controller in the Struts architecture, and the ActionServlet component binds the task parameters in the production process file and assembles them into a task package Model and sends it to the execution engine.

[0031] Optionally, the execution engine receives the task package sent by the task scheduling module, parses the production process information and task parameters in the task package, and sends them to the execution controller. The execution controller loads the received production process node information and task parameters, starts execution from the first node of the production process, calls the corresponding software robot to process the satellite image according to the task parameters of this node, and then returns the execution status of this step to the execution controller. After receiving it, the execution controller determines whether it is completed normally. If it is normal, it controls the production process to enter the next node until the production process is completed. If it is abnormal, it feedbacks to the task scheduling module that the production task terminates.

[0032] The advantages of the present invention compared with the prior art are as follows: The embodiment of the present invention designs a process-based and configurable automatic production mode of satellite images based on software robots, splits various processing steps required for satellite image production into independent functional software robots one by one, encapsulates input and output parameters according to the data transmission protocol defined by the system, and the production task process design adopts a graphical method. According to the use of various series of functional software robots in the satellite image production process, the production process is designed by arranging and combining, and task parameters are bound according to needs to form a task instance. The spatial data engine automatically schedules hardware resources and software robots to complete satellite image production to meet the needs of satellite image automatic production and improve the production efficiency of satellite images. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 FIG. is a schematic structural diagram of an automatic satellite image production system based on a spatial software robot provided by an embodiment of the present invention;

[0034] Figure 2 FIG. is a schematic diagram of the composition of a system structure provided by an embodiment of the present invention;

[0035] Figure 3 FIG. is a schematic diagram of an application example provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0036] Referring to Figure 1 , which shows a schematic structural diagram of an automatic satellite image production system based on a spatial software robot provided by an embodiment of the present invention. As Figure 1 shown, the system may include: a robot center 110, a robot design module 120, a vector spatialization module 130, and a spatial data engine 140, where

[0037] the robot center 110 may be configured to be encapsulated and developed according to the operations required in the satellite image production process;

[0038] The robot design module 120 can be configured to select appropriate functional software robots from the robot center according to the satellite production operation process for combination and setting;

[0039] The vector space conversion module 130 can be configured to perform decompression on the original satellite image, extract the original image for original image extraction work, as well as perform mathematical modeling work such as metadata creation and vector extraction of the image range, and coordinate conversion and projection conversion processing work;

[0040] The spatial data engine 140 can be configured to automatically retrieve the production process matching the task from the robot design module according to the production task initiated by the operator, bind the task parameters input by the operator to the production process to create a task instance, and call each functional software robot according to the designed production process to process the original image data to produce satellite images.

[0041] In a specific implementation manner of the present invention, the robot center includes: a preprocessing software robot, a rectification and fusion software robot, an inspection and analysis software robot, a uniform light and color software robot, a mosaicking and finishing software robot, a data cropping software robot, a metadata production software robot, and a data slicing software robot. Among them,

[0042] The preprocessing software robot is configured to decompress the original image and extract the original metadata to form a list of images to be processed;

[0043] The rectification and fusion software robot is configured to rectify and perform band fusion on a single-frame original image;

[0044] The inspection and analysis software robot is configured to inspect and analyze the rectified single-frame image;

[0045] The uniform light and color software robot is configured to adjust the exposure and color of a single-frame image;

[0046] The mosaicking and finishing software robot is configured to splice multiple-frame images and perform transitional finishing processing on the splicing edges;

[0047] The data cropping software robot is configured to crop the mosaicked image according to the standard map sheet;

[0048] The metadata production robot is configured to generate new metadata from the metadata extracted from the original image according to the standard map sheet;

[0049] The data slicing robot is configured to process the result data into slice data for publishable image services.

[0050] In another specific implementation manner of the present invention, the robot design module includes: a start / end node sub-module, a robot node sub-module, a control node sub-module, an interaction node sub-module, and a production process management sub-module, where,

[0051] The start / end node sub-module is configured to control the start and end of the production process;

[0052] The robot node sub-module is configured to select functional software robots of each series in the robot center;

[0053] The control node sub-module is configured to control the production process;

[0054] The interaction node sub-module is configured to pop up prompt information for the operator to confirm for the production steps that require manual intervention, and rebind some task parameters during the execution of the production process;

[0055] The production process management sub-module is configured to perform new operation, modification operation, and deletion operation on the production process.

[0056] In another specific implementation manner of the present invention, the spatial data engine includes: a parameter binding sub-module, a task scheduling sub-module, and an execution engine, where,

[0057] The parameter binding sub-module is configured to automatically retrieve a production process file matching the task from the robot design module according to the production task initiated by the operator, and bind the task parameters input by the operator to create a task instance of the production process;

[0058] The task scheduling sub-module is configured to detect the total number of task instances in the current system, determine whether there are idle resources to execute the task instance, and if there are idle resources, automatically split the task into multiple sub-tasks that can be executed in parallel, and package the original image data to be produced and the task instance into a task package and send it to the execution engine configured for each hardware resource;

[0059] The execution engine is configured to receive and load the task package sent by the task scheduling module, parse the production process information and task parameters in the task package, start executing from the first node of the production process, and call the corresponding software robot to process the satellite image according to the task parameters of the node.

[0060] In another specific implementation manner of the present invention, the parameter binding sub-module is implemented based on the MEF framework, and is specifically configured to use the Container in the MEF framework as the main body for loading the production process, map the parameters that the production task needs to transfer to the Container through the Import interface of the MEF framework, assign parameters and set attributes to each node of the production process through the Export interface of MEF, and save the selected nodes, assigned parameters and attribute values of the designed production process as a task instance.

[0061] In another specific implementation manner of the present invention, the task scheduling module uses the ActionServlet component in the Struts architecture as the controller. According to the struts-config.xml configuration file that describes the corresponding relationship between the model, view, and controller in the Struts architecture, it retrieves the production process file that matches the assigned task from the robot design module, and the ActionServlet component binds the task parameters in the production process file and assembles them into a task package Model and sends it to the execution engine.

[0062] In another specific implementation manner of the present invention, the execution engine receives the task package sent by the task scheduling module, parses the production process information and task parameters in the task package, and sends them to the execution controller. The execution controller loads the received production process node information and task parameters, starts executing from the first node of the production process, calls the corresponding software robot to process the satellite image according to the task parameters of this node, and then returns the execution status of this step to the execution controller. After receiving it, the execution controller judges whether it is completed normally. If it is normal, it controls the production process to enter the next node until the production process is completed. If it is abnormal, it feedbacks to the task scheduling module that the production task is terminated.

[0063] For the above system provided by the embodiments of the present invention, it can be combined with Figure 2 for the following detailed description.

[0064] As Figure 2 shown, the invention mainly includes a robot center, a robot design module, and a spatial data engine.

[0065] The robot center includes a series of functional software robots such as preprocessing, correction and fusion, inspection and analysis, equalization of light and color, mosaicking and finishing, data cropping, metadata production, and data slicing. Each series of functional software robots is encapsulated and developed according to the operations required in the satellite image production process. The preprocessing software robot mainly completes the decompression of the original image, the extraction of the original metadata, and the formation of a list of images to be processed; the correction and fusion software robot mainly completes the correction and band fusion of a single original image; the inspection and analysis software robot mainly completes the inspection and analysis of the corrected single image; the equalization of light and color software robot mainly completes the adjustment of the exposure and color of a single image; the mosaicking and finishing software robot mainly completes the splicing of multiple images and the transitional finishing of the splicing edges; the data cropping software robot is mainly responsible for cropping the mosaicked image according to the standard map sheets; the metadata production robot is used to regenerate new metadata from the metadata extracted from the original image according to the standard map sheets; the data slicing robot is used to process the resulting data into slice data for publishable image services. Each functional software robot includes input parameters and output parameters, and the functional software robots in the robot center are the basis for the design and production process of the robot design module.

[0066] The robot design module selects appropriate functional software robots from the robot center according to the satellite production operation process for combination and setting, including start / end nodes, robot nodes, control nodes, interaction nodes, and production process management, and designs the production process in a graphical way; the start / end nodes are used to control the start and end of the production process; the robot nodes can select various series of functional software robots in the robot center, and each node needs to be configured with input parameters and output parameters. When designing the production process, fixed values can be input for each parameter, or parameter variables can be used, and binding is performed according to the task parameters input by the operator when starting the production process; the control nodes include branch nodes, loop nodes, and nested nodes. The branch node is used to judge the process direction according to conditions, the loop node is used to execute certain steps in a loop, and the nested node is used to nest a sub-process in the main process; the interaction nodes include dialog nodes and parameter modification nodes. The dialog node is used to pop up a prompt message for the operator to confirm in the steps that require manual intervention, and the parameter modification node is used to rebind some task parameters during the execution of the production process; the production process management includes the creation, modification, and deletion of the process.

[0067] The vector space module includes original image extraction, mathematical modeling, coordinate transformation, and projection transformation. This module mainly completes the work of original image extraction such as decompressing and extracting the original satellite images. Since different types of original satellite image data are organized differently, which affects the automatic reading of the data production process, decompression and extraction are required and the data should be organized according to unified rules. Mathematical modeling work such as metadata creation and vector extraction of image range preliminarily creates the specifications of the image production results, and coordinate transformation and projection transformation unify the specifications of the image production results.

[0068] The spatial data engine includes parameter binding, task scheduling, and execution engine. The parameter binding module automatically retrieves the production process file matching the task from the robot design module according to the production task initiated by the operator, and binds the task parameters input by the operator to the production process to create a task instance; this module is implemented based on the MEF framework, uses the Container in the MEF framework as the main body for loading the production process, maps the parameters that the production task needs to transfer to the Container through the Import interface of the MEF framework, and assigns parameters and sets attributes to each node of the production process through the Export interface of MEF; then saves each node, assigned parameters, and attribute values selected by the designed production process as a task instance.

[0069] The task scheduling module detects the total number of task instances in the current system, determines whether there are idle resources to execute this task instance. If there are idle resources, it automatically splits this task into multiple sub-tasks that can be executed in parallel, and packages the original image data to be produced and the task instance into a task package and sends it to the execution engine configured for each hardware resource; this module uses the ActionServlet component in the Struts architecture as the controller. According to the struts-config.xml configuration file that describes the corresponding relationship between the model, view, and controller in the Struts architecture, it retrieves the production process file matching the assigned task from the robot design module, and the ActionServlet component binds the task parameters in the production process file and assembles them into a task package Model and sends it to the execution engine.

[0070] The execution engine is implemented using a microservices architecture. Each execution engine is a controller container that dynamically assigns an execution controller to each production process instance to control the execution of the process. A single server or workstation can run multiple production processes, and the executions of these processes are independent of each other and do not affect each other. The controller container receives and loads the task package sent by the task scheduling module, parses the production process information and task parameters in the task package, and sends them to the execution controller. The execution controller loads the received production process node information and task parameters, starts executing from the first node of the production process, calls the corresponding software robot to process the satellite image according to the task parameters of this node, and then returns the execution status of this step to the execution controller. After receiving it, the execution controller determines whether it is completed normally. If it is normal, it controls the production process to enter the next node until the production process is completed. If it is abnormal, it feeds back to the task scheduling module that the production task terminates.

[0071] As Figure 3 shown, in an embodiment of the present invention, after receiving the production task operation of the operator, the spatial data engine first performs vector spatialization processing such as original image extraction, mathematical modeling, coordinate transformation, and projection transformation to form the to-be-processed image data with unified data specifications and standardized attribute fields.

[0072] Then, it starts the parameter binding program. The parameter binding module retrieves the production process file matching the production task from the robot design module, and binds the task parameters set by the operator into the production process to form a task instance.

[0073] The task scheduling module allocates hardware resources and the to-be-processed original satellite images for the task instance, packages them into a task package and sends it to the execution engine. The execution engine loads and parses the task package. First, it calls the preprocessing function software robot to perform operations such as decompressing the original satellite image and extracting the original metadata to form a list of to-be-processed original images. Then, it takes the first scene of the original satellite image and sequentially calls the function software robots such as rectification and fusion, inspection and analysis, and equalization of illumination and color for processing. After the processing is completed, it queries the list of to-be-processed images to determine whether all are completed. If not, it takes the next scene of the original image and continues the processing. If completed, it continues to sequentially call the software robots such as mosaicking and finishing, data cropping, and metadata production to complete the subsequent operations. After the production process reaches the end node, it ends the production task and feeds back to the task scheduling module.

[0074] The specific embodiments described in this application can enable those skilled in the art to understand this application more comprehensively, but do not limit this application in any way. Therefore, those skilled in the art should understand that they still make modifications to this application or equivalent replacements; and all technical solutions and their improvements that do not depart from the spirit and technical essence of this application should be covered by the protection scope of the patent of this application.

[0075] The content not described in detail in the specification of the present invention belongs to the well-known technology of those skilled in the art.

Claims

1. A satellite image automated production system based on a space software robot, characterized in that, The system includes: a robot center, a robot design module, a vector spatialization module, and a spatial data engine. Among them, the robot center is configured to be encapsulated and developed according to the operations required in the satellite image production process; the robot design module is configured to select appropriate functional software robots from the robot center according to the satellite production operation process for combination and setting; the vector spatialization module is configured to decompress the original satellite image, extract the original image from the image, perform metadata creation, vector extraction mathematical modeling of the image range, and coordinate transformation and projection transformation processing; the spatial data engine is configured to automatically retrieve the production process matching the task from the robot design module according to the production task initiated by the operator, bind the task parameters input by the operator to the production process to create a task instance, and call each functional software robot according to the designed production process to process the original image data and produce satellite images.

2. The system according to claim 1, wherein The robot center includes: a preprocessing software robot, a rectification and fusion software robot, an inspection and analysis software robot, a color equalization software robot, a mosaicking and finishing software robot, a data cropping software robot, a metadata production software robot, and a data slicing software robot. Among them, the preprocessing software robot is configured to decompress the original image and extract the original metadata to form a list of images to be processed; the rectification and fusion software robot is configured to rectify and perform band fusion on a single-scene original image; the inspection and analysis software robot is configured to inspect and analyze the rectified single-scene image; the color equalization software robot is configured to adjust the exposure and color of a single-scene image; the mosaicking and finishing software robot is configured to splice multiple-scene images and perform transitional finishing processing on the splicing edges; the data cropping software robot is configured to crop the mosaicked image according to the standard map sheet; the metadata production software robot is configured to generate new metadata from the metadata extracted from the original image according to the standard map sheet; the data slicing software robot is configured to process the resulting data into slice data for publishable image services.

3. The system according to claim 1, wherein The robot design module includes: a start / end node sub-module, a robot node sub-module, a control node sub-module, an interaction node sub-module, and a production process management sub-module. Among them, the start / end node sub-module is configured to control the start and end of the production process; the robot node sub-module is configured to select various series of functional software robots in the robot center; the control node sub-module is configured to control the production process; the interaction node sub-module is configured to pop up prompt information for the operator to confirm for production steps requiring manual intervention, and rebind some task parameters during the execution of the production process; the production process management sub-module is configured to perform new, modification, and deletion operations on the production process.

4. The system according to claim 1, wherein The spatial data engine includes: a parameter binding sub-module, a task scheduling sub-module, and an execution engine. Among them, The parameter binding sub-module is configured to automatically retrieve a production process file that matches the task from the robot design module according to a production task initiated by an operator, and bind the task parameters input by the operator to the production process to create a task instance; The task scheduling sub-module is configured to detect the total number of task instances in the current system, determine whether there are idle resources to execute the task instance, and if there are idle resources, automatically split the task into multiple sub-tasks that can be executed in parallel, and package the original image data to be produced and the task instance into a task package and send it to the execution engine configured for each hardware resource; The execution engine is configured to receive and load the task package sent by the task scheduling module, parse the production process information and task parameters in the task package, start executing from the first node of the production process, and call the corresponding software robot to process the satellite image according to the task parameters of this node.

5. The system according to claim 4, characterized in that The parameter binding sub-module is implemented based on the MEF framework. Specifically, it is configured to use the Container in the MEF framework as the main body for loading the production process, map the parameters that the production task needs to transfer to the Container through the Import interface of the MEF framework, perform parameter assignment and attribute setting for each node of the production process through the Export interface of MEF, and save each node, the assigned parameters and attribute values selected by the designed production process as a task instance.

6. The system according to claim 4, wherein The task scheduling module uses the ActionServlet component in the Struts architecture as the controller. According to the struts-config.xml configuration file that describes the model, view, and controller correspondence in the Struts architecture, it retrieves the production process file that matches the assigned task from the robot design module, and the ActionServlet component binds the task parameters in the production process file and assembles them into a task package Model and sends it to the execution engine.

7. The system according to claim 4, characterized in that The execution engine receives the task package sent by the task scheduling module, parses the production process information and task parameters in the task package, and sends them to the execution controller. The execution controller loads the received production process node information and task parameters, starts executing from the first node of the production process, calls the corresponding software robot to process the satellite image according to the task parameters of this node, and then returns the execution status of this step to the execution controller. After receiving it, the execution controller determines whether it is completed normally. If it is normal, it controls the production process to enter the next node until the production process is completed. If it is abnormal, it feedbacks to the task scheduling module that the production task is terminated.

Images