Visual space science experiment data analysis pipeline method

Abstract

The embodiment of the invention provides a visual space science experiment data analysis assembly line method. The technical problem that space science experiment data analysis is poor in convenience can be solved. The method comprises the following steps: in response to input data selection operation, algorithm model selection operation and visual component selection operation, generating data nodes, model nodes and visual nodes; performing parameter configuration on the data nodes, the model nodes and the visual nodes, and performing assembly line connection operation to generate a flow chart; based on the flow chart, receiving an input job submission operation; analyzing the flow chart to obtain an analysis result; obtaining a job scheduling result based on the analysis result; based on the job scheduling result, an algorithm mirror image deployment package corresponding to the algorithm model selection operation and file data corresponding to the data selection operation are called; performing algorithm analysis on the file data according to the algorithm mirror image deployment package to obtain an analysis result; and performing visual display. And multi-process, interactive and quantitative data analysis is realized.

Description

Technical Field

[0001] The present application relates to the technical field of data analysis, and in particular to a visualized pipeline method for analyzing space science experiment data. Background Art

[0002] The online analysis and application research of space science multi-task data plays a decisive role in improving the research and utilization level of massive and multi-source space science experimental data, promoting the output of results and knowledge discovery, and better exerting the application benefits of space science experimental data.

[0003] At present, computer technology can be used to visualize electronic information on computer devices, thereby realizing human-computer interactive data analysis research.

[0004] However, the user interface in the related technology does not support interactive parameter configuration of each node. Users cannot directly view the analysis results visually in the current user interface. Users need to manually manage the algorithm modules in the server, and the scheduling efficiency of algorithm analysis jobs needs to be improved. These make user data analysis inconvenient. Summary of the invention

[0005] The embodiment of the present application provides a visual pipeline method for space science experiment data analysis, which can solve the technical problem of poor convenience in space science experiment data analysis.

[0006] To achieve the above objectives, the embodiments of the present application adopt the following technical solutions:

[0007] In a first aspect, an embodiment of the present application provides a visualized space science experiment data analysis pipeline method, which includes: generating data nodes, model nodes and visualization nodes in response to input data selection operations, algorithm model selection operations and visualization component selection operations; the data nodes, model nodes and visualization nodes are located on the same user interface; the data nodes, model nodes and visualization nodes are parameterized, and pipeline connection operations are performed to generate a flow chart; wherein the output of the data node is connected to the input of the model node, or the output of the data node is connected to the input of the visualization node, and the output of the model node is connected to the visualization node The method comprises the steps of: receiving input of a node; receiving input of a job submission operation based on a flowchart; parsing the flowchart to obtain parsing results; using the parsing results to characterize the dependency relationships among data nodes, model nodes, and visualization nodes and the sorting results of data flows; obtaining job scheduling results based on the parsing results; retrieving the algorithm image deployment package corresponding to the algorithm model selection operation and the file data corresponding to the data selection operation based on the job scheduling results; the file data includes video data and image data; performing algorithmic analysis on the file data according to the algorithm image deployment package to obtain analysis results; and visually displaying the analysis results; wherein the analysis results include images, videos, character strings, two-dimensional charts, and / or three-dimensional charts.

[0008] Based on the above description of the visualized space science experiment data analysis pipeline method provided in the embodiment of the present application, it can be seen that the visualized space science experiment data analysis pipeline method includes interactively completing node selection and layout, node relationship connection, data and algorithm configuration, analysis progress monitoring and analysis result visualization in the same user interface, so that the entire data algorithm analysis process forms a visual, seamless whole, thereby improving the intuitiveness of user data analysis.

[0009] In addition, the embodiments of the present application implement multi-process pipeline analysis operations and their automated management, and support parallel orchestration of multiple algorithm analysis processes. After the orchestration is completed, the system automatically completes the establishment of dependency relationships for algorithm analysis operations, calls the analysis algorithm to run the operation, and collects analysis progress, analysis results, and operation logs, greatly simplifying the user's operation complexity and reducing the system's learning cost and usage threshold.

[0010] Furthermore, the embodiment of the present application realizes the display of raw data and analysis results in the same interface. When the analysis job is completed, the user can view the visualization of raw data and analysis results synchronously in the current user interface, intuitively compare and display the differences and characteristics between the analysis results and the raw data, and improve the convenience of user data analysis.

[0011] In addition, the embodiments of the present application support any programming language and framework through automated deployment of algorithm models and automated management of analysis tasks, allowing algorithm developers to focus on the development of the algorithm itself without having to pay attention to the underlying software and hardware resources of the system, as well as the operation scheduling of the analysis algorithm, which helps to improve the scalability of the system and improve the utilization rate of the underlying software and hardware resources.

[0012] In a feasible implementation of the first aspect, the visualized space science experiment data analysis pipeline method further includes: based on web page embedding, setting the data node, the model node and the visualization node on the same user interface.

[0013] In a feasible implementation of the first aspect, the visualized space science experiment data analysis pipeline method also includes: the pipeline connection operation includes a multi-process pipeline node connection operation; the multi-process pipeline node connection operation includes a node connection method based on a directed acyclic graph structure, connecting the isolated data nodes, the model nodes and the visualization nodes into a complete algorithm analysis pipeline.

[0014] In a feasible implementation of the first aspect, the visualized space science experiment data analysis pipeline method also includes: based on the flowchart, verifying whether there is an isolated node; if the isolated node does not exist, then based on the flowchart, verifying whether there are missing input edges; if there are no missing input edges, then based on the flowchart, verifying whether the data types match; if the data types match, then based on the flowchart, verifying whether the required parameters are filled in; if the required parameters are all filled in, then based on the flowchart, verifying whether there is a loop; if the loop does not exist, then based on the flowchart, receiving the input job submission operation.

[0015] In a feasible implementation of the first aspect, the visualized space science experiment data analysis pipeline method also includes: obtaining multiple jobs based on the flowchart; for the jobs having the first relationship, sending them one by one in a serial form to the analysis algorithm on the computing instance.

[0016] In a feasible implementation of the first aspect, the visualized space science experiment data analysis pipeline method further includes: obtaining multiple jobs based on the flowchart; and for the jobs having the second relationship, sending them to the analysis algorithm on the computing instance in parallel.

[0017] In a feasible implementation of the first aspect, the visualized space science experiment data analysis pipeline method also includes: the algorithm image deployment package is used to cooperate in generating an execution container; the execution container includes a processor, memory, graphics processor resources and a mounted file directory.

[0018] In a feasible implementation of the first aspect, the visualized space science experiment data analysis pipeline method also includes: presetting an image template; reading the image template based on the algorithm configuration file, copying the code and data of the algorithm model into the image, automatically installing dependent third-party libraries, and building the algorithm image deployment package.

[0019] In a second aspect, an embodiment of the present application provides a visualized space science experiment data analysis pipeline system, which includes: at least one processor; a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method provided in the first aspect.

[0020] In a third aspect, an embodiment of the present application provides a computer-readable medium having computer program instructions stored thereon, and the computer program instructions can be executed by a processor to implement the method provided in the first aspect. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 A schematic diagram of the structure of a visual space science experiment data analysis pipeline system provided in an embodiment of the present application;

[0022] Figure 2 A schematic diagram of the structure of a visual space science experiment data analysis pipeline system provided in an embodiment of the present application;

[0023] Figure 3 A flow chart of a visual space science experiment data analysis pipeline method provided in an embodiment of the present application;

[0024] Figure 4 A schematic diagram of a data selection node in a visualized space science experiment data analysis pipeline method provided in an embodiment of the present application;

[0025] Figure 5 A schematic diagram of an algorithm model node in a visualized space science experiment data analysis pipeline method provided in an embodiment of the present application;

[0026] Figure 6 A schematic diagram of a data visualization node in a visualized space science experiment data analysis pipeline method provided in an embodiment of the present application;

[0027] Figure 7 A log diagram of the automated deployment of an algorithm model in a visualized pipeline method for space science experiment data analysis provided in an embodiment of the present application;

[0028] Figure 8A schematic diagram of a calculation example in a visualized space science experiment data analysis pipeline method provided in an embodiment of the present application;

[0029] Fig. 9 A flow chart of a visual space science experiment data analysis pipeline method provided in an embodiment of the present application;

[0030] Fig.10 A visual overall interface of a flowchart of a visual space science experiment data analysis pipeline method provided in an embodiment of the present application. DETAILED DESCRIPTION

[0031] The technical solutions in the embodiments of the present invention will be described below in conjunction with the accompanying drawings in the embodiments of the present invention. In the description of the embodiments of the present invention, unless otherwise specified, "multiple" refers to two or more than two. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single items or plural items. For example, at least one of a, b, or c can represent: a, b, c, ab, ac, bc, or abc, where a, b, and c can be single or multiple.

[0032] In addition, in order to facilitate the clear description of the technical solutions of the embodiments of the present invention, in the embodiments of the present invention, the words "first", "second" and the like are used to distinguish the same items or similar items with substantially the same functions and effects. Those skilled in the art will understand that the words "first", "second" and the like do not limit the quantity and execution order, and the words "first", "second" and the like do not necessarily limit the difference. At the same time, in the embodiments of the present invention, the words "exemplary" or "for example" are used to indicate examples, illustrations or explanations. Any embodiment or design described as "exemplary" or "for example" in the embodiments of the present invention should not be interpreted as being more preferred or more advantageous than other embodiments or design. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a concrete way for easy understanding.

[0033] The principles and features of the present application are described below. The examples given are only used to explain the present application and are not used to limit the scope of the present application.

[0034] Visualization refers to the technology of displaying data, information, and knowledge in an intuitive and easy-to-understand way through visualization tools such as charts, graphs, and maps, which helps people better understand the development process of events and discover the laws contained in the data. The user interaction interface based on visualization can enable users to find the required information or functions more quickly and reduce the time required to complete tasks. Completing data selection, data analysis configuration, and result visualization in the same user interface is a necessary way to achieve efficient and convenient data analysis, and is also the development trend of data visualization analysis software today.

[0035] Automated deployment is a process that automatically deploys applications from a development environment to a production or other environment through software tools. This process reduces the need for manual operations, thereby reducing the risk of human error and improving deployment efficiency and frequency. Automated deployment usually includes steps such as code building, testing, configuration management, and application release to ensure that software updates can be deployed to the target system quickly and reliably.

[0036] Dynamic scheduling is a method of adjusting and optimizing workload distribution based on resource availability and job demand at runtime. This method is particularly suitable for environments where there are dependencies between jobs and resources are limited. It can adjust the scheduling policy based on the real-time system status and resource usage to cope with unforeseen changes. Since dynamic scheduling makes decisions based on real-time information, task priorities and execution order can be adjusted during operation. By dynamically reallocating jobs to the most suitable resources, the overall performance of the system can be maximized. If a job fails or lacks necessary resources, dynamic scheduling can reschedule the affected jobs. Therefore, dynamic scheduling is often used to manage the execution of compute-intensive tasks, especially in cloud computing, data analysis and other fields.

[0037] The embodiment of the present application provides a visual pipeline method for space science experiment data analysis, which is suitable for engineering application scenarios of data types such as manned space science videos and images. By designing a flowchart user interface, automatic deployment of algorithm models and dynamic scheduling of analysis operations, multi-process, interactive, intuitive and quantitative space science experiment data algorithm analysis is achieved.

[0038] The embodiment of the present application provides a visualized space science experiment data analysis pipeline system, which can execute the visualized space science experiment data analysis pipeline method provided by the embodiment of the present application. Figure 1 A structural schematic diagram of a visual space science experiment data analysis pipeline system provided in an embodiment of the present application.

[0039] like Figure 1As shown, the visualized space science experiment data analysis pipeline system includes at least one processor 011 and a memory 012 communicatively connected to the at least one processor; wherein the memory 012 stores instructions executable by the at least one processor 011, and the instructions are executed by the at least one processor 011 so that the at least one processor 011 can execute the visualized space science experiment data analysis pipeline method provided in the embodiment of the present application.

[0040] In some embodiments, the visualized space science experiment data analysis pipeline system may be an electronic device provided with a display screen.

[0041] The display screen can be used to display the user interface and realize visualization and human-computer interaction functions. For example, display images, videos, etc. The display screen includes a display panel. The display panel can be a liquid crystal display (LCD), an organic light-emitting diode (OLED), an active-matrix organic light emitting diode or an active-matrix organic light emitting diode (AMOLED), a flexible light-emitting diode (FLED), Miniled, MicroLed, Micro-oLed, quantum dot light emitting diodes (QLED), etc. In some embodiments, the electronic device may include 1 or N display screens, where N is an integer greater than 1.

[0042] Exemplarily, the electronic devices in the embodiments of the present application may include, but are not limited to, folding screen devices (e.g., folding screen mobile phones, folding screen computers, etc.), tablet computers, PDAs, netbooks, personal computers (PCs), personal digital assistants (PDAs), portable multimedia players (PMPs), augmented reality (AR) / virtual reality (VR) devices, televisions, car computers, etc. Alternatively, the electronic device may also be an electronic device with a display screen of another type or structure. The embodiments of the present application do not limit the specific device form of the electronic device.

[0043] like Figure 2As shown, in some embodiments, the visualized space science experiment data analysis pipeline system also includes a flowchart user interface module, a distributed database and file system, an algorithm model automatic deployment module, an analysis job dynamic scheduling module, and a computing instance middleware module.

[0044] In one implementation, the flowchart user interface module includes a data selection component, an algorithm model component, and a data visualization component.

[0045] In one implementation, the flowchart user interface module also includes a multi-process pipeline node connection component, a node parameter configuration component, and a multi-process pipeline node layout component.

[0046] Figure 3 The present invention provides a flow chart of a visual space science experiment data analysis pipeline method. Figure 3 As shown, in some embodiments, the visualized space science experiment data analysis pipeline method includes the following steps:

[0047] S1, generating data nodes, model nodes and visualization nodes in response to input data selection operations, algorithm model selection operations and visualization component selection operations.

[0048] Data nodes, model nodes, and visualization nodes are located on the same user interface.

[0049] The data selection operation refers to selecting video data or image data as data items to be analyzed by the algorithm. In some embodiments, the video data or image data can be pre-stored in a distributed database and a file system. When performing the data selection operation, it can be achieved by selecting the file name of the video data or image data.

[0050] like Figure 4 As shown, in one implementation, the user can select "Original Video Data.mp4" by clicking the selection box "Open Data Panel" on the user interface to implement the user's data selection operation. It is understandable that the data selection operation can also be implemented through non-contact operations such as user eye movement, language, or gestures. The specific form of the data selection operation is not limited in this application.

[0051] In one implementation, in response to an input data selection operation, a Figure 4 The data node shown. This data node component has only one output. This output is used to connect to the input of a model node or visualization node.

[0052] In some embodiments, a data node component is dragged onto the canvas, and a data item is selected through the data menu on the data node. After saving, the data node represents the data item. The data node has only an output interface.

[0053] The algorithm model selection operation refers to selecting a model. In some embodiments, the data related to the algorithm can be pre-stored in the algorithm model automatic deployment module. When performing the model selection operation, it can be implemented by selecting the name of the model. Figure 5 As shown, in one implementation, the user clicks the selection box "Universal Semantic Segmentation Model" on the user interface to implement the user's algorithm model selection operation. It is understandable that the algorithm model selection operation can also be implemented through non-contact operations such as the user's eye movement, language, or gestures. This application does not limit the specific form of the algorithm model selection operation.

[0054] In one implementation, in response to an input algorithm model selection operation, a Figure 5 The model node shown. The model node component has one or more inputs, and can have one or more outputs. Among them, one output can be connected to multiple downstream inputs, and one input can only be connected to one upstream output, that is, the input can only have one source, and the output can be copied multiple times for sharing by multiple downstream nodes. In one implementation, select a model in the algorithm model list and drag it to the canvas to form an algorithm model node. The model node has input and output interfaces, as well as a parameter form.

[0055] In some embodiments, data nodes, model nodes, and visualization nodes are uniformly displayed on a display screen. Visualization can be achieved through a data visualization component. In one implementation, the data visualization component is implemented in a web page (WEB) embedded form, supports image comparison visualization, video visualization and other data visualization displays directly in the current user interface, and also supports opening the data visualization view in the visualization node in a separate window in full screen mode.

[0056] S2, configure the parameters of the data nodes, model nodes and visualization nodes, and perform pipeline connection operations to generate a flow chart.

[0057] Parameter configuration operations include setting algorithm model parameters. Parameter configuration supports multiple data types. For example, the data type can be a number. For example, an integer or a floating point number. Visually displayed as a number input box, select the number input box, and enter a number in the number input box. For another example, the data type can be a string. Visually displayed as a character input box, select the character input box, and enter a string in the character input box. For another example, the data type can be Boolean. Visually displayed as a check box. For another example, the data type can be a single choice. Visually displayed as a drop-down box. Figure 5As shown, in one implementation, the user clicks the input box "iou" on the user interface and then enters the value "0.6" to implement the parameter configuration operation of the parameter intersection-over-union ratio threshold. It is understandable that the parameter configuration operation can also be implemented through non-contact operations such as user eye movement, language, or gestures. This application does not limit the specific form of the parameter configuration operation.

[0058] In one implementation, each node parameter configuration module is used to perform parameter configuration on a specific node component.

[0059] Pipeline connection operations, such as Fig.10 As shown, in some embodiments, a connection line is dragged between nodes (data nodes, model nodes, and visualization nodes). When connecting, the output interface of a node is connected to the input interface of the next node. In this way, all node connections form a directed acyclic graph. For example, the output of a data node is connected to the input of a model node, or the output of a data node is connected to the input of a visualization node.

[0060] In some embodiments, the pipeline connection operation can be implemented through a flowchart component. The user generates a flowchart by connecting various data nodes, model nodes, or visualization nodes on the display screen. After connecting to any qualified node, the node component has one or more inputs. The flowchart view can be scaled and moved with the flowchart as a whole.

[0061] In some embodiments, the pipeline connection operation may be a multi-process pipeline node connection. In one implementation, the multi-process pipeline node connection connects isolated nodes into a complete algorithm analysis pipeline in the form of lines, and supports parallel layout and relational connection of nodes based on a node connection method of a directed acyclic graph structure.

[0062] It can be understood that each process pipeline in the flowchart can be regarded as a job (also called an algorithm analysis job).

[0063] In some embodiments, the flowchart and form are saved to the database in JSON format, and the information includes node information, node position and edge information. The node information includes the data corresponding to the node, the algorithm model, the form content, and the visualization component ID.

[0064] The node position includes the horizontal coordinate and vertical coordinate of the node on the canvas. The edge information includes the source node and target node of each edge, and the source interface and target interface of each edge. In one implementation, the multi-process pipeline layout uses an automatic layout algorithm to determine the horizontal coordinate and vertical coordinate of each node in the user interface so that they are neatly arranged without overlap.

[0065] S3, based on the flowchart, receives the input job submission operation.

[0066] The operation submission includes the user inputting by clicking the "Submit" button on the display screen after the flow chart is generated. It is understandable that the data selection operation can also be implemented through non-contact operations such as user eye movement, language, or gestures. The specific form of the data selection operation is not limited in this application.

[0067] In some embodiments, the job submission operation can be input through a job submission interface. The job submission interface is implemented by developing an algorithm according to an interface specification.

[0068] In some embodiments, in addition to the job submission interface, a job query interface and a job cancellation interface may also be provided.

[0069] In some embodiments, when executing step S3, the visualized space science experiment data analysis pipeline method further includes:

[0070] S31, based on the flow chart, verify the correctness of its parameters and topological relationships.

[0071] In one implementation, when executing step S31, the visualized space science experiment data analysis pipeline method further includes:

[0072] S311, based on the flowchart, verify whether there is an isolated node.

[0073] S312: If there is no isolated node, verify whether there is any missing input edge based on the flowchart.

[0074] S313: If there is no missing input edge, verify whether the data type matches based on the flowchart.

[0075] S314: If the data types match, verify whether the required parameters are filled in based on the flowchart.

[0076] S315: If all required parameters have been filled in, verify whether there is a loop based on the flow chart.

[0077] A loop, also called a cycle, is a concept in graph theory. If task x must be completed before task y: x→y, and y→z, and z→x. Then a cycle will be formed: x→y→z→x. A valid flowchart should not have loops, i.e. a directed acyclic graph.

[0078] S316: If there is no loop, based on the flowchart, receive an input job submission operation.

[0079] S4, analyze the flowchart and obtain the analysis results.

[0080] The parsing results are used to characterize the dependencies between data nodes, model nodes, and visualization nodes and the sorting results of data flows.

[0081] Parse the flowchart, including analyzing the algorithm analysis flowchart arranged by the user, extracting the input data, algorithm model, configuration parameters and topological relationship, and sorting the nodes in the flowchart according to their dependency relationship and data flow.

[0082] In some embodiments, the flowchart information is handed over to the analysis job dynamic scheduling module for flowchart analysis.

[0083] S5, obtaining the job scheduling result based on the analysis result.

[0084] According to the analysis results, including the node dependencies, the algorithm analysis job is sent to the corresponding computing instance in a serial or parallel manner and runs. At the same time, the job execution progress is monitored and the job is scheduled. In this way, the job scheduling result is obtained. Figure 8 An example of a computational instance is given.

[0085] In some embodiments, all algorithm model nodes in the flowchart are topologically sorted according to the node information and edge information to determine the sequence and dependency of each analysis job.

[0086] In one implementation, for jobs with a first relationship (i.e., a dependency relationship), the analysis algorithms on the computing instances are sent serially one by one. After the analysis algorithm completes the job, the system will update the process progress and execute subsequent jobs until all analysis jobs corresponding to all process nodes are completed.

[0087] In another implementation, for jobs with a second relationship (i.e., no dependency), the analysis algorithm is sent to the computing instance in parallel. After the analysis algorithm completes the job, the system will update the process progress and execute subsequent jobs until all analysis jobs corresponding to all process nodes are completed.

[0088] In some embodiments, when executing step S5, the visualized space science experiment data analysis pipeline method further includes:

[0089] S51, displaying the progress of each job according to the job scheduling result.

[0090] In some embodiments, progress monitoring can display the progress of the analysis job in the algorithm model node in real time. Exemplarily, the progress of the job completion is expressed in the form of a percentage of the completed amount to the total amount.

[0091] In some embodiments, the progress of each job is uploaded to a distributed database and file system.

[0092] S6, based on the job scheduling result, retrieve the algorithm image deployment package corresponding to the algorithm model selection operation and the file data corresponding to the data selection operation.

[0093] The file data includes video data and image data. In some embodiments, the data to be analyzed is downloaded from a distributed database and a file system.

[0094] The video data is a data file stored in a video format, such as mp4, mpeg, avi, etc. The image data format supports png, jpg, bmp, etc., which are data files stored in an image compression package format, such as rar, zip, etc.

[0095] The algorithm image deployment package is used to cooperate in generating an execution container. The algorithm image deployment package can be pre-stored in a visualized space science experiment data analysis pipeline system. In some embodiments, the algorithm image deployment package is pre-stored in an algorithm model automation deployment module.

[0096] In some embodiments, the interface and parameter definition files are written according to the system interface specification and packaged together with the algorithm program as the algorithm program to be deployed. In this way, the algorithm image deployment package is obtained.

[0097] The execution container includes processor (CPU), memory, graphics processor (GPU) resources and mounted file directories, and has the functions of generating, starting, stopping, restarting and deleting. The algorithm container can also be scheduled to execute and cancel analysis tasks. In some embodiments, the computing instance middleware executes container engine (Docker, a virtualized container engine at the operating system level) commands to start, stop, restart, delete, read, etc. of the container. Figure 7 Log and other operations shown.

[0098] The construction of the algorithm image deployment package is described in detail below.

[0099] In some embodiments, before executing step S1, the visualized space science experiment data analysis pipeline method further includes:

[0100] S611, preset image template.

[0101] An image template library is stored in the algorithm model automatic deployment module. The image template library includes multiple image templates. In some embodiments, based on the Docker container engine, according to different hardware and software environments, more than ten basic container image templates composed of different hardware environments and software environments can be preset, such as: Python language program image template that requires GPU, Python language program image template that does not require GPU, Java language program image template that requires GPU, Java language program image template that does not require GPU, etc. In one implementation method, the selected basic image is copied to the algorithm program directory, the algorithm program is built as a Docker image according to the process described in the basic image, and the installation script is run. Then the operating environment is automatically installed for it according to the environment configuration file in the algorithm program. After the image is built, the image is exported as a compressed file for deployment.

[0102] Image templates provide operating systems, system-level libraries, and corresponding installation scripts.

[0103] S612: Build an algorithm image deployment package based on the image template.

[0104] Image template, used for building algorithm image deployment packages. In some embodiments, the image template is read from the image template library according to the algorithm configuration file, the code and data of the algorithm model are copied to the image template, and the third-party programs and components that the algorithm model depends on are automatically installed (such as: MMCV for computer vision, NumPy for numerical calculations), and finally the algorithm image is generated and packaged and stored in the specified location of the server. It is understandable that the requirements of the third-party programs on which the algorithm model program depends can be described by the environment file. When building the image, the environment file can be read to install it through the network.

[0105] In order to ensure the accuracy of constructing the algorithm image deployment package, in some embodiments, before presetting the image template, the algorithm interface information and data type are tested to see if they meet the preset standards. Steps S611 and S612 can be executed only after the test is successful. Fig. 9 As shown, in one implementation, the correctness of the algorithm program is verified by reading the interface and parameter definitions in the algorithm program.

[0106] In some embodiments, the algorithm program is uploaded to the system, and the name, description and other information are entered, and the required image template and the computing instance to be deployed are selected. After the algorithm is uploaded, the visual space science experiment data analysis pipeline system will automatically execute the testing, packaging and deployment processes. In one implementation, the exported image is copied to the computing instance and imported into Docker to form a container. The input and output data directories are mounted for the container, and then the container is started to complete the deployment.

[0107] S7, performing algorithm analysis on the file data according to the algorithm image deployment package to obtain analysis results.

[0108] The analysis results store all the data generated by the algorithm, which can be directly visualized in this system and also supports data download.

[0109] S8, visualize the analysis results.

[0110] In some embodiments, analysis results are uploaded to a distributed database and file system.

[0111] In some embodiments, the visualization display can be implemented through a data visualization component. In one implementation, the data visualization component is implemented in the form of an embedded web page, supporting data visualization displays such as image comparison visualization and video visualization directly in the current user interface. The data visualization view can be scaled and moved with the overall flowchart, and it also supports opening the data visualization view in a separate window in full screen mode.

[0112] In some embodiments, a visualization component is selected from the visualization component list and dragged onto the canvas to form a visualization node. The visualization node has only an input interface. Fig. 9 As shown in the figure, integrate the developed visualization components into the system. The integrated visualization components will appear in the visualization component list.

[0113] In order to meet the needs of users, such as Fig. 9 As shown, in one implementation, declaring a visualization component requires input of a data type and quantity.

[0114] In some embodiments, an interactive data visualization component is developed, including an image visualization function, a video playback function, a string (such as JSON) viewing function, a two-dimensional chart visualization function, and a three-dimensional chart visualization function. The image visualization function can be used to view a series of images. If there are algorithm analysis results, they can be displayed synchronously with the original data. The video playback function can be used to play videos. The JSON viewing function displays the properties and values ​​of the JSON file in a tree structure. The two-dimensional chart visualization function displays data in the form of scattered points, broken lines, parallel coordinate axes, etc. on a two-dimensional coordinate system. The three-dimensional chart visualization function displays data in the form of scattered points, broken lines, surfaces, etc. on a three-dimensional coordinate system.

[0115] The embodiment of the present application realizes a unified interactive visual user interface. By designing a flowchart-style user interface, node views containing different information such as editable parameters or viewing status are designed for data selection, algorithm configuration, and analysis result visualization, and interactive node layout, relationship connection, and parameter setting are supported to integrate all information into the same page. The design determines the source and flow of data based on the connection relationship between nodes, and determines the running order of analysis jobs.

[0116] The embodiment of the present application realizes a multi-process data algorithm analysis pipeline by designing a node connection method of a directed acyclic graph structure and a dynamic scheduling technology for analysis operations. The design uses a directed acyclic graph structure to carry out parallel layout and relationship connection of nodes. Each node connection branch forms an algorithm analysis pipeline to realize a multi-process algorithm analysis pipeline, forming an algorithm analysis flowchart that integrates data selection, algorithm configuration, and analysis result visualization. When the algorithm analysis flowchart is set, the system will sort the nodes in the flowchart according to their dependencies and data flow, and create an algorithm analysis job. The dynamic scheduling technology for analysis operations will send the algorithm analysis job to the corresponding computing instance in a serial or parallel manner according to the dependency of the nodes and run it. At the same time, it monitors the progress of job execution and performs job scheduling until all analysis jobs corresponding to all nodes are completed.

[0117] The embodiment of the present application realizes visualization of analysis results in the same interface by designing a data visualization view embedded in a WEB page, and embedding the view directly into the current user interface. The view can accept multiple data sources through node connections, and synchronously display the original data and analysis results in the embedded visualization view. The visualization view can be scaled and moved with the overall flowchart, and also supports opening the data visualization view in a separate window in full-screen mode.

[0118] The embodiment of the present application realizes the automatic deployment of the algorithm model. By designing the algorithm model automatic deployment module, the algorithm model interface declaration standard is clarified. When the algorithm development is completed, the system first tests whether the algorithm model runs normally and verifies whether the algorithm model interface declaration file meets the standard. If the algorithm model passes the test, the system automatically installs the dependent components for it and packages it into a Docker image. Subsequently, the packaged image is automatically sent to the computing instance, which is responsible for loading the algorithm model image and providing analysis services to the outside world.

[0119] Based on the same application concept, a visualized space science experiment data analysis pipeline system is also provided in an embodiment of the present application. The method corresponding to the visualized space science experiment data analysis pipeline system may be the visualized space science experiment data analysis pipeline method in the aforementioned embodiment, and its principle of solving the problem is similar to that of the method. The visualized space science experiment data analysis pipeline system provided in an embodiment of the present application includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the methods and / or technical solutions of the aforementioned multiple embodiments of the present application.

[0120] Another embodiment of the present application further provides a computer-readable storage medium having computer program instructions stored thereon, wherein the computer program instructions can be executed by a processor to implement the methods and / or technical solutions of any one or more embodiments of the present application described above.

[0121] Specifically, the present embodiment may adopt any combination of one or more computer-readable media. The computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination thereof. More specific examples (non-exhaustive list) of computer-readable storage media include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In this document, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in combination with an instruction execution system, device, or device.

[0122] Computer readable signal media may include a data signal propagated in baseband or as part of a carrier wave, which carries a computer readable program code. Such propagated data signals may take a variety of forms, including, but not limited to, electromagnetic signals, optical signals, or any suitable combination of the above. Computer readable signal media may also be any computer readable medium other than a computer readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0123] Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

[0124] Computer program code for performing the operations of the present application may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0125] The flow chart or block diagram in the accompanying drawings shows the possible architecture, function and operation of the equipment, method and computer program product according to various embodiments of the present application. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some implementations as replacements, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated system for hardware that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0126] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0127] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or page components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0128] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0129] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0130] The above-mentioned integrated unit implemented in the form of a software functional unit can be stored in a computer-readable storage medium. The above-mentioned software functional unit is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) or a processor to perform some steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), disk or optical disk and other media that can store program codes.

[0131] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

[0132] In addition, it is clear that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units or devices stated in a device claim can also be implemented by one unit or device through software or hardware. The words first, second, etc. are used to indicate names, and do not indicate any particular order.

Claims

1. A visual pipeline method for space science experiment data analysis, characterized in that: include: In response to input data selection operations, algorithm model selection operations, and visualization component selection operations, data nodes, model nodes, and visualization nodes are generated; The data node, the model node and the visualization node are located on the same user interface; Parameter configuration is performed on the data node, the model node and the visualization node, and pipeline connection operation is performed to generate a flow chart; wherein the output of the data node is connected to the input of the model node, or the output of the data node is connected to the input of the visualization node, and the output of the model node is connected to the input of the visualization node; Based on the flow chart, receiving an input job submission operation; Parsing the flowchart to obtain parsing results; the parsing results are used to characterize the dependency relationship between the data nodes, the model nodes and the visualization nodes and the sorting result of the data flow; Based on the analysis result, a job scheduling result is obtained; Based on the job scheduling result, the algorithm image deployment package corresponding to the algorithm model selection operation and the file data corresponding to the data selection operation are retrieved; the file data includes video data and image data; According to the algorithm image deployment package, performing algorithm analysis on the file data to obtain an analysis result; Visually display the analysis results; wherein the analysis results include images, videos, character strings, two-dimensional charts and / or three-dimensional charts.

2. The visualized space science experiment data analysis pipeline method according to claim 1 is characterized in that: The visualized space science experiment data analysis pipeline method also includes: based on web page embedding, setting the data node, the model node and the visualization node on the same user interface.

3. The visualized space science experiment data analysis pipeline method according to claim 1 or 2, characterized in that: The visualized space science experiment data analysis pipeline method also includes: The pipeline connection operation includes a multi-process pipeline node connection operation; the multi-process pipeline node connection operation includes a node connection method based on a directed acyclic graph structure, connecting the isolated data nodes, the model nodes and the visualization nodes into a complete algorithm analysis pipeline.

4. The visualized space science experiment data analysis pipeline method according to claim 1 or 2, characterized in that: The visualized space science experiment data analysis pipeline method also includes: Based on the flow chart, verify whether there is an isolated node; If the isolated node does not exist, verifying whether there is a missing input edge based on the flow chart; If the missing input edge does not exist, verify whether the data types match based on the flow chart; If the data types match, verify whether the required parameters are filled in based on the flow chart; If the required parameters are all filled in, verify whether there is a loop based on the flow chart; If the loop does not exist, then based on the flow chart, the job submission operation is received as input.

5. The visualized space science experiment data analysis pipeline method according to claim 1 or 2, characterized in that: The visualized space science experiment data analysis pipeline method also includes: Based on the flow chart, a plurality of jobs are obtained; The jobs having the first relationship are sent to the analysis algorithms on the computing instances one by one in a serial manner.

6. The visualized space science experiment data analysis pipeline method according to claim 1 or 2, characterized in that: The visualized space science experiment data analysis pipeline method also includes: Based on the flow chart, a plurality of jobs are obtained; For jobs with the second relationship, the analysis algorithm is sent to the computing instance in parallel.

7. The visualized space science experiment data analysis pipeline method according to claim 1 or 2, characterized in that: The visualized space science experiment data analysis pipeline method also includes: The algorithm image deployment package is used to cooperate in generating an execution container; the execution container includes a processor, memory, graphics processor resources and a mount file directory.

8. The visualized space science experiment data analysis pipeline method according to claim 7, characterized in that: The visualized space science experiment data analysis pipeline method also includes: Preset image templates; According to the algorithm configuration file, the image template is read, the code and data of the algorithm model are copied to the image, the dependent third-party library is automatically installed, and the algorithm image deployment package is built.

9. A visual space science experiment data analysis pipeline system, characterized in that: include: at least one processor; a memory communicatively coupled to the at least one processor; The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method according to any one of claims 1 to 8.

10. A computer readable medium having computer program instructions stored thereon, wherein the computer program instructions can be executed by a processor to implement the method according to any one of claims 1 to 8.

Images