Remote sensing image data management method and device and medium

Abstract

The invention discloses a remote sensing image data management method and apparatus, and a medium. The method comprises the steps of obtaining image metadata; defining each image picture in the image metadata as an entity node in a blood relationship graph; wherein the entity nodes comprise image nodes and slice nodes; distributing a unique data ID (Identity) for the entity node to serve as DNA (Deoxyribonucleic Acid) of the image picture; determining a data flow direction between the entity nodes; and constructing a blood relationship graph according to the data ID and the data flow direction. Therefore, the unique ID is allocated to the image metadata, namely, the unique data DNA is generated, so that the blood relationship graph about the remote sensing image is constructed, the image picture corresponding to each entity node can be quickly and efficiently checked and traced based on the blood relationship graph, the deficiencies and redundant links in the image data processing flow can be timely found, and the processing efficiency of the remote sensing image is improved. While efficient management of remote sensing images is realized, the quality of image data is improved, and more accurate image data support is provided for various fields.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a remote sensing image data management method, device and medium. Background Art

[0002] With the continuous development of satellite technology, the amount of remote sensing image data has exploded. In the actual use of remote sensing images, it is necessary to perform multi-level data processing on the original remote sensing images to obtain the final usable image slices.

[0003] In the current image data processing chain, due to the diversity and complexity of processing procedures and methods, problems such as redundancy in data processing, data missing, and data inconsistency often occur. These problems directly affect the accuracy and reliability of remote sensing image data, thus causing great trouble to the subsequent remote sensing image data analysis and application.

[0004] If we want to promptly discover deficiencies and redundancies in the data processing process and identify potential problems such as data duplication, missing data and errors, then it is an urgent problem to quickly, efficiently and accurately trace the source of image slices, data processing process and application direction. Summary of the invention

[0005] In view of this, one aspect of the present application provides a remote sensing image data management method, the method comprising:

[0006] Get image metadata;

[0007] Define each image picture in the image metadata as an entity node in the blood relationship graph; wherein the entity node includes an image node and a slice node;

[0008] Assigning a unique data ID to the physical node to serve as the DNA of the image;

[0009] Determining the data flow direction between the entity nodes;

[0010] The blood relationship graph is constructed according to the data ID and the data flow direction.

[0011] Optionally, the data ID includes an image ID and a slice ID; and the assigning a unique data ID to the entity node includes:

[0012] Determine a first blood relationship between the image nodes and a second blood relationship between the image nodes and the slice nodes;

[0013] According to the first blood relationship, assigning a unique image ID to each image node;

[0014] According to the second blood relationship and the image ID, a unique slice ID is allocated to each slice node; wherein the slice ID includes a part of the image ID.

[0015] Optionally, if the slice node includes a parent slice node and a child slice node, and the slice ID includes a parent slice ID and a child slice ID; then, according to the second blood relationship and the image ID, a unique slice ID is allocated to each slice node, including:

[0016] Determine a third blood relationship between the parent slice node and the child slice node;

[0017] According to the second blood relationship and the image ID, a unique parent slice ID is allocated to the parent slice node; wherein the parent slice ID includes a part of the image ID;

[0018] According to the third blood relationship, a unique sub-slice ID is assigned to the sub-slice node; wherein the sub-slice ID includes a part of the parent slice ID.

[0019] Optionally, determining the data flow direction between the entity nodes includes:

[0020] identifying an image processing task in the image metadata;

[0021] According to the image processing task, determining whether the generation of one node between any two of the entity nodes depends on the other node, so as to determine the dependency relationship between any two of the entity nodes;

[0022] The data flow direction is determined according to the dependency relationship.

[0023] Optionally, constructing the blood relationship graph according to the data ID and the data flow direction includes:

[0024] Extracting image task information and node attribute information of the entity node from the image metadata; wherein the image task information at least includes image processing timestamp, image processing method, and image production information; and the node attribute information at least includes node ownership information, node storage path, and node application type;

[0025] Injecting the node attribute information into the corresponding entity node;

[0026] According to the data flow direction, adding directed edges between the entity nodes;

[0027] The corresponding image task information is added to the directed edges to construct the blood relationship graph.

[0028] Optionally, the remote sensing image data management method further includes:

[0029] When receiving an image tracking instruction, parsing the image tracking instruction to determine a starting node for tracking;

[0030] Among all entity nodes in the blood relationship graph, locate the target entity node corresponding to the start node;

[0031] The tracking result of the target entity node is sent to a terminal; wherein the tracking result at least includes all reachable paths between the target entity nodes.

[0032] Optionally, the remote sensing image data management method further includes:

[0033] When an update instruction to update the current blood relationship graph is obtained;

[0034] Determine whether there is a target image picture that is the same as the newly added image picture in the current blood relationship graph;

[0035] If so, determine whether the quality of the newly added image is higher than that of the target image; if so, replace the newly added image with the target image; if not, discard the newly added image;

[0036] If not present, the newly added image picture is added to the current blood relationship diagram.

[0037] Another aspect of the present application provides a remote sensing image data management device, the device comprising:

[0038] An image metadata acquisition module is used to acquire image metadata;

[0039] An entity node definition module, used to define each image picture in the image metadata as an entity node in the blood relationship graph; wherein the entity node includes an image node and a slice node;

[0040] An ID allocation module, used to allocate a unique data ID to the entity node to serve as the DNA of the image;

[0041] A data flow direction determination module, used to determine the data flow direction between the entity nodes;

[0042] A construction module is used to construct the blood relationship diagram according to the data ID and the data flow direction.

[0043] Another aspect of the present application provides a remote sensing image data management device, including a memory and a processor, wherein the memory stores a computer program that can be executed on the processor, and when the processor executes the program, the steps of the remote sensing image data management method are implemented.

[0044] Another aspect of the present application provides a computer-readable storage medium having a computer program stored thereon, and when the program is executed by a processor, the steps of the remote sensing image data management method are implemented.

[0045] The present application provides a remote sensing image data management method, device and medium, which have the following beneficial effects: thus, by assigning a unique ID to the image metadata, that is, generating a unique data DNA, a blood relationship diagram of the remote sensing image is constructed. Based on the blood relationship diagram, it is possible to quickly, efficiently and accurately view and trace the image pictures corresponding to each entity node, timely discover deficiencies and redundant links in the image data processing process, and realize efficient management of remote sensing images while improving the quality of remote sensing image data, providing more accurate image data support for various fields. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 A flowchart of a remote sensing image data management method provided in an embodiment of the present application;

[0047] Figure 2 A visualization schematic diagram of a blood relationship diagram of a remote sensing image provided in an embodiment of the present application;

[0048] Figure 3 A flowchart of a remote sensing image data management method provided by another embodiment of the present application;

[0049] Figure 4 A schematic diagram of the structure of a remote sensing image data management device provided in an embodiment of the present application;

[0050] Figure 5 A schematic diagram of the structure of a remote sensing image data management device provided in another embodiment of the present application.

[0051] The reference numerals are as follows: 50 is a memory, 51 is a processor, 52 is a display screen, 53 is an input / output interface, 54 is a communication interface, 55 is a power supply, 56 is a communication bus, 501 is a computer program, 502 is an operating system, and 503 is data. DETAILED DESCRIPTION

[0052] The terms used in this application are for the purpose of describing specific embodiments only and are not intended to limit this application. The singular forms of "a", "said" and "the" used in this application and the appended claims are also intended to include plural forms unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used herein refers to and includes any or all possible combinations of one or more associated listed items.

[0053] It should be understood that although the terms first, second, third, etc. may be used in the present application to describe various information, these information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of the present application, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Depending on the context, the word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining".

[0054] Figure 1 A flowchart of a remote sensing image data management method provided in an embodiment of the present application is shown in FIG. Figure 1 As shown, the method includes:

[0055] S10: Obtain image metadata;

[0056] In a specific embodiment, the image metadata may be collected from a database storing remote sensing image data by using a data collection tool. Specifically, as an optional embodiment, the image metadata may be collected from the database by using a data collection tool Talend.

[0057] It is worth noting that, for the database storing remote sensing image data, one database may be used for storage, or different databases may be used to store different types of remote sensing image data, which is not limited in this application. In an optional embodiment, different image metadata may be obtained from multiple different databases such as a MySQL database, a PostGIS database, a MinIO database, and a MongoDB database.

[0058] In addition, it should be noted that the acquired image metadata includes but is not limited to image pictures, image attribute information and image mission information, and the image attribute information includes but is not limited to the image name, the name of the satellite that acquired the image, the sensor type, and the shooting time. Image mission information includes but is not limited to image processing timestamp, image processing method, and image production information. Among them, the image processing method includes but is not limited to image processing mode, image processing algorithm, and algorithm parameters. Image production information includes but is not limited to production timestamp, production unit, and production person in charge.

[0059] It is understandable that the initial remote sensing image acquired by the satellite is a large image whose content cannot be directly recognized by the naked eye. It needs to undergo multiple image processing before it can become an image that can be directly recognized by the naked eye. Among them, image processing includes but is not limited to cropping the image. Therefore, the image metadata obtained in step S10 includes unsegmented image pictures and sliced ​​image pictures.

[0060] In an optional embodiment, data on unsegmented images, image attribute information and image task information corresponding to the unsegmented images may be obtained from a PostGIS database. Data on sliced ​​images, image attribute information and image task information corresponding to the sliced ​​images may be obtained from a MongoDB database.

[0061] In addition, in another optional embodiment, the image metadata may also include a data storage path, an image application type, etc. The image application type refers to the specific application type, usage scenario, etc. of the image. Correspondingly, in an optional embodiment, the data storage path may be obtained from the MinIO database, and the image application type may be obtained from the MySQL database.

[0062] In order to improve the overall quality of remote sensing images, in an optional embodiment, after the image metadata is acquired through step S10, the remote sensing image data management method provided in the embodiment of the present application further includes preprocessing the image metadata.

[0063] Specifically, the image metadata is filtered to filter out image data with low resolution and poor quality. Then, the filtered image metadata is classified to determine the management scope of the current image metadata. When classifying, it can be divided according to whether the image picture is segmented, that is, classified into unsegmented image pictures and sliced ​​image pictures. It can also be divided according to the application type of the image picture. This application does not limit the classification method of the image.

[0064] In addition, preprocessing the image metadata may also include converting the data into a unified format. It is understandable that different databases may store data in different formats. In order to facilitate unified management of data, after obtaining the image metadata stored in each database, all data are converted into a unified format.

[0065] S11: defining each image picture in the image metadata as an entity node in a blood relationship graph; wherein the entity node includes an image node and a slice node;

[0066] Furthermore, in order to construct a kinship graph of remote sensing images, it is necessary to define each node in the kinship graph. Specifically, each image picture in the image metadata is defined as an entity node in the kinship graph. It can be understood that the image picture includes an unsegmented image picture and a sliced ​​image picture, so the entity node includes an image node and a sliced ​​node. Among them, the image node is the node of the unsegmented image picture, and the slice node is the node of the sliced ​​image picture.

[0067] S12: assigning a unique data ID to the entity node as the DNA of the image;

[0068] S13: Determine the data flow direction between the entity nodes;

[0069] S14: Construct a blood relationship graph based on the data ID and data flow direction.

[0070] In order to locate and track each entity node in the future, it is necessary to assign a unique data ID to each entity node, so as to set a unique DNA for each image. Determine the data flow between different entity nodes, that is, determine the dependency between different entity nodes. In order to build a blood relationship diagram of remote sensing images based on entity nodes, data IDs of entity nodes and data flow directions.

[0071] In an optional embodiment, in addition to assigning unique data IDs to different entity nodes, different node names are set for different entity nodes, so that each entity node can be located by different node names, that is, the image picture can be located. In an optional embodiment, the node name can be set according to the direction of the data flow. Specifically, different methods such as text and numbers can be used. The specific form of the node name is not limited in this application.

[0072] In order to facilitate users to directly view the constructed blood relationship diagram and trace each entity node in the blood relationship diagram, in an optional embodiment, the blood relationship diagram can be visualized. Specifically, the visualization tool Tableau can be used to display the blood relationship between each image picture, and through an interactive interface, users can analyze the slice blood relationship by entering information such as data ID, and obtain image pictures, image attribute information and image task information, etc.

[0073] In an optional embodiment, the blood relationship diagram can be displayed in a graphical manner, and the present application does not limit the manner of displaying the blood relationship. Figure 2 A visualization diagram of a blood relationship diagram of a remote sensing image provided in an embodiment of the present application is shown below for ease of understanding. Figure 2 Provide explanation.

[0074] It should be noted that Figure 2 What is shown is a schematic diagram of some blood relationships of remote sensing images, such as Figure 2 As shown, the entity nodes in the blood relationship graph include multiple image nodes 20 and multiple slice nodes 21. Different node names are set for different image nodes. Specifically, the multiple image nodes 20 include image node L0, image node L1, image node L2 and image node L3. And the n slice nodes 21 are also set with different slice node names, for example, slice node L30.

[0075] exist Figure 2 In the blood relationship diagram shown, different entity nodes are set with corresponding unique data IDs, which can be used to trace different images later. In addition, the data flow direction is determined between different entity nodes, and the following is constructed based on the data ID and data flow direction: Figure 2 Schematic diagram of blood relationship shown.

[0076] It should be noted that, in an optional embodiment, different entity nodes can be represented by circles, and when the entity nodes corresponding to different circles are clicked, the information corresponding to the entity nodes can be popped up. Figure 2 As shown, different entity nodes can be represented by rectangles, and each rectangle is labeled with a corresponding node name, and different entity nodes are connected to a rectangle including corresponding node information. This application does not limit the form of visualization of the blood relationship diagram.

[0077] In an optional embodiment, the visualization tool Tableau can be used to display the lineage diagram of remote sensing images. Through an interactive interface, users are allowed to analyze and obtain the lineage relationship between entity nodes by entering data IDs. At the same time, the source, processing method and direction of the data can be obtained and displayed in a graphical manner. Users can easily obtain images of interest and understand how these images are cut out from the original remote sensing images and obtain the final results after various processing steps.

[0078] In another optional embodiment, the remote sensing image data management method provided in this application can also provide users with image image blood relationship query and analysis functions. Specifically, the user enters information such as data ID to query blood relationship, and automatically analyzes the blood relationship, and generates a corresponding analysis report to provide to the user. In an optional embodiment, the user can also set customized analysis and report generation according to actual business needs.

[0079] Therefore, the remote sensing image data management method provided in the embodiment of the present application constructs a lineage diagram of remote sensing images by assigning a unique ID to the image metadata, that is, generating a unique data DNA. Based on the lineage diagram, it is possible to quickly, efficiently and accurately view and trace the image pictures corresponding to each entity node, timely discover deficiencies and redundant links in the image data processing process, and achieve efficient management of remote sensing images while improving the quality of remote sensing image data, providing more accurate image data support for various fields.

[0080] Figure 3 This is a flow chart of a remote sensing image data management method provided by another embodiment of the present application. In an optional embodiment, the data ID includes an image ID and a slice ID. Figure 3 As shown, a unique data ID is assigned to the entity node, including:

[0081] S30: Determine a first blood relationship between image nodes and a second blood relationship between image nodes and slice nodes;

[0082] It can be understood that the image picture includes an unsegmented image picture and a sliced ​​image picture, and correspondingly, the entity node includes an image node and a slice node, wherein the image node is a node of the unsegmented image picture, and the slice node is a node of the sliced ​​image picture. When allocating a unique data ID to all entity nodes, it is necessary to allocate it to both the image node and the slice node, so correspondingly, the data ID includes the image ID and the slice ID.

[0083] The data ID is used as a unique identifier of the entity node, that is, as a unique identifier of each image, so that any image and its related information can be traced back according to the constructed blood relationship graph.

[0084] In the specific embodiment of assigning a unique data ID to each entity node, in order to more intuitively observe the relationship between different entity nodes when visualizing the blood relationship graph, as an optional embodiment, the first blood relationship between the image nodes and the second blood relationship between the image nodes and the slice nodes are first determined.

[0085] It is understandable that different entity nodes represent different images, and there may be dependencies between different images. Specifically, the generation of one image depends on another image. Figure 2As shown in the figure, the image picture corresponding to the image node L1 is obtained by image processing the image picture corresponding to the image node L0. Therefore, the image node L0 can be regarded as the parent node of the image node L1, and the image node L1 can be regarded as the child node of the image node L0. That is, there is a first blood relationship between the image node L0 and the image node L0. Similarly, it can be determined that Figure 2 In the blood relationship diagram shown, there is a second blood relationship between any slice node in the set consisting of the image node L3 and the slice node 21.

[0086] S31: assigning a unique image ID to each image node according to the first blood relationship;

[0087] Further, after determining the first blood relationship between the image nodes, a unique image ID is classified for each image node according to the parent node and the child node in the first blood relationship. In an optional embodiment, a number can be used as the image ID. In this embodiment, the parent image ID of the parent node is smaller than the child image ID corresponding to the child node, and the parent image ID and the child image ID are set according to the numerical order. For example, Figure 2 In the blood relationship diagram shown, image node L0 is the parent node, and the corresponding parent image ID can be set to 001; image node L1 is the child node, and the corresponding child image ID is 002.

[0088] It should be noted that, in another optional embodiment, the data ID may also be any combination of text, English, numbers, etc. Table 1 is a definition information table of an entity node of an image picture provided in an embodiment of the present application. For ease of understanding, the following will be described in conjunction with Table 1.

[0089] For example, as shown in Table 1, image node L0 is the parent node of image node L1, image node L1 is the parent node of image node L2, and image node L2 is the parent node of image node L3. Therefore, the image IDs corresponding to each image node are set according to the numerical order. In addition, English is added to the image ID to indicate that the current image ID is an image node, not a slice ID. For example, as shown in Table 1, the image ID of image node L0 is ImgNode_L0_001, where "ImgNode_L0" is used to indicate that the current node is an image node and specifically image node L0, and "001" can be used to indicate the relationship with other image nodes. The image IDs of other image nodes are similarly shown in Table 1.

[0090] Of course, in another optional embodiment, a randomly generated sequence can be used as the data ID, as long as the randomly generated sequence is guaranteed to be a unique sequence. In summary, the present application does not limit the representation form of the data ID.

[0091] Table 1 Definition information table of entity nodes of image pictures

[0092] Serial number Entity Node Node Name Data ID 1 Image Node Image node L0 ImgNode_L0_001 2 Image Node Image node L1 ImgNode_L1_002 3 Image Node Image node L2 ImgNode_L2_003 4 Image Node Image Node L3 ImgNode_L3_004 5 Slice Node Slice node L1_001 SliceNode_L1_002_001 6 Slice Node Slice node L2_001 SliceNode_L2_003_001 7 Slice Node Slice node L2_002 SliceNode_L2_003_002 8 Slice Node Slice node L2_002_S1 SliceNode_L2_003_002_S1 …… …… …… ……

[0093] S32: Allocate a unique slice ID to each slice node according to the second blood relationship and the image ID; wherein the slice ID includes a part of the image ID.

[0094] Similarly, a corresponding slice ID needs to be assigned to each slice node. Specifically, in an optional embodiment, since the slice image corresponding to the slice node has a dependency relationship with the unsegmented image corresponding to the image node, the slice image can be obtained after the unsegmented image is segmented.

[0095] Therefore, when assigning slice IDs, in order to more intuitively observe the relationship between image nodes and slice nodes when visualizing the blood relationship diagram, a unique slice ID can be assigned to each slice node based on the second blood relationship and the image ID; wherein the slice ID includes a part of the image ID.

[0096] For example, as shown in Table 1, the slice node L1_001 is obtained by splitting the image node L1, so the slice ID of the slice node L1_001 is SliceNode_L1_002_001, which includes "L1_002" in the image ID corresponding to the image node L1. It can be understood that "SliceNode" is used to indicate that the current node is a slice node, and "L1_002" is used to indicate that the current node has a blood relationship with the node whose ID includes "L1_002".

[0097] In fact, in an optional embodiment, the node names between image nodes and between image nodes and slice nodes can correspond to the relationship between data IDs. That is, the slice node name includes a part of the image node name. Between image nodes, the image node name corresponding to the parent node and the image node name corresponding to the child node are set according to the numerical order. In this way, the dependency relationship between each entity node can be viewed and traced more intuitively.

[0098] Based on the above embodiment, as an optional embodiment, if the slice node includes a parent slice node and a child slice node, the slice ID includes a parent slice ID and a child slice ID. Figure 3 As shown, according to the second blood relationship and the image ID, a unique slice ID is assigned to each slice node, including:

[0099] S320: Determine the third blood relationship between the parent slice node and the child slice node;

[0100] It is understandable that in the actual application of remote sensing images, high resolution is required for specific areas such as cities, while relatively low resolution is required for areas such as oceans. Therefore, for areas such as cities, the image may need to be segmented more finely. At this time, the parent node of the slice may not be an image node, but a slice node, that is, the slice may be segmented from a larger slice, that is, the slice node includes a parent slice node and a child slice node.

[0101] Therefore, in order to realize the tracing of slice nodes, it is necessary to assign unique IDs to both the parent slice node and the child slice node, that is, the slice ID includes the parent slice ID and the child slice ID. In a specific embodiment, it is necessary to first determine the third blood relationship between the parent slice node and the child slice node, which is similar to determining the first blood relationship between image nodes. Please refer to the description of the above embodiment, which will not be repeated here.

[0102] S321: Allocate a unique parent slice ID for the parent slice node according to the second blood relationship and the image ID; wherein the parent slice ID includes a part of the image ID;

[0103] It can be understood that when a slice node includes a parent slice node and a child slice node, the image picture corresponding to the parent slice node should be obtained by splitting the image picture corresponding to one of the image nodes. Therefore, when assigning a unique parent slice ID to the parent slice node, in an optional embodiment, it can be assigned based on the second blood relationship between the image node and the slice node and the image ID of the image node.

[0104] Specifically, the parent slice ID includes a part of the image ID. For example, as shown in Table 1, the slice node with the node name slice node L2_001 is a slice of the image node L2, so the parent slice ID of slice node L2_001 is SliceNode_L2_003_001, which includes "L2_003" in the image ID of image node L2. Similarly, slice node L2_002 is also a slice of image node L2, so the parent slice ID of slice node L2_002 is SliceNode_L2_003_002, which includes "L2_003" in the image ID of image node L2.

[0105] S322: According to the third blood relationship, a unique sub-slice ID is assigned to the sub-slice node; wherein the sub-slice ID includes a part of the parent slice ID.

[0106] Further, after the parent slice ID is generated, a unique sub-slice ID of the sub-slice node can be allocated according to the third blood relationship and the parent slice ID. Specifically, the sub-slice ID includes a portion of the parent slice ID. It can be understood that the image corresponding to the parent slice node is segmented to obtain the image corresponding to the sub-slice node. Therefore, in an optional embodiment, a portion of the parent slice ID can be used as one of the components of the sub-slice ID, thereby intuitively determining the relationship between the parent slice node and the sub-slice node.

[0107] For example, as shown in Table 1, the slice node named slice node L2_002_S1 is a child slice of slice node L2_002. Therefore, slice node L2_002 is the parent slice node of slice node L2_002_S1, and the child slice ID of slice node L2_002_S1 is SliceNode_L2_003_002_S1, that is, including "L2_003_002" in the parent slice ID.

[0108] It can be understood that in the sub-slice ID "SliceNode_L2_003_002_S1", "SliceNode" is used to indicate that the current ID is a slice ID, and "L2_003_002" is used to indicate that the current node has a blood relationship with the node whose ID includes "L2_003_002".

[0109] In an optional embodiment, determining the data flow direction between entity nodes includes:

[0110] Identify image processing tasks in image metadata;

[0111] According to the image processing task, determine whether the generation of one node between any two entity nodes depends on another node, so as to determine the dependency relationship between any two entity nodes;

[0112] Determine the data flow direction based on the dependencies.

[0113] In a specific embodiment, when constructing a blood relationship graph, after determining each entity node, it is also necessary to determine the dependency relationship between different entity nodes, that is, determine the data flow direction between different entity nodes, so as to trace the image processing relationship between different entity nodes.

[0114] Specifically, each image processing task is first identified from the acquired image metadata, and based on the image processing task, it is determined whether the generation of one node between any two entity nodes depends on the generation of another node, that is, whether the generation of an image corresponding to one node depends on the generation of an image corresponding to another node. If there is a dependency relationship, it can be determined that there is a dependency relationship between the two entity nodes, and further, the data flow direction can be determined based on the dependency relationship.

[0115] For example, Figure 2 In the blood relationship diagram shown, image node L1 is the parent node of image node L2, that is, the generation of image node L2 needs to rely on image node L1. Therefore, in a specific embodiment, there is a data flow direction between image node L1 and image node L2. Specifically, image node L1 flows to image node L2 to generate image node L2 after image processing of image node L1.

[0116] Based on the above embodiment, as an optional embodiment, a blood relationship graph is constructed according to the data ID and the data flow direction, including:

[0117] Extract image task information and node attribute information of entity nodes from image metadata; wherein the image task information at least includes image processing timestamp, image processing method, and image production information; the node attribute information at least includes node ownership information, node storage path, and node application type;

[0118] Inject node attribute information into the corresponding entity node;

[0119] Add directed edges between entity nodes according to the direction of data flow;

[0120] Add the corresponding image task information to the directed edges to build a blood relationship graph.

[0121] After defining each entity node and determining the data flow direction between the entity nodes, the blood relationship graph can be constructed. Specifically, in order to realize the traceability of images and related information, when constructing the blood relationship graph, all the information corresponding to each image needs to be injected into the relationship graph.

[0122] Therefore, the image task information and node attribute information of the entity node are extracted from the image metadata. Among them, the image task information refers to the relevant information of the image processing of each image picture, specifically including but not limited to the image processing timestamp, image processing method, and image production information. The image processing timestamp refers to the specific time when the image picture is processed. The image processing method includes but is not limited to the image processing mode, image processing algorithm, and algorithm parameters. The image processing method includes but is not limited to image correction, enhancement, fusion, cropping, and slicing. Image production information includes but is not limited to the production timestamp, production unit, and production person in charge.

[0123] The node attribute information of an entity node refers to the image attribute information of the image picture corresponding to the entity node, wherein the node attribute information includes image attribute information and slice attribute information, and the node attribute information at least includes node ownership information, node storage path and node application type. The node storage path includes different storage paths such as image pictures at each level, thumbnails, slices, sample sets, and topics.

[0124] The node ownership information is not limited to the image size, image name, satellite name for obtaining the image, sensor type, and shooting time. Figure 2 As shown, the node application types include but are not limited to application sample sets, thematic analysis, geographic information research, etc. It should be noted that the image pictures need to be processed for corresponding applications before being used as application sample sets, and the application processing includes but is not limited to ground feature classification, graphic interpretation, target detection and semantic segmentation.

[0125] Furthermore, the node attribute information is injected into the corresponding entity node, that is, the node attribute information is added to the corresponding entity node. It should be noted that, when visualizing, the node attribute information can be as follows: Figure 2 What is shown is direct display. Of course, in order to improve aesthetics, the node attribute information can be hidden. When the corresponding entity node is clicked, the corresponding node attribute information will be displayed. This application does not limit this.

[0126] While injecting the node attribute information, according to the data flow direction determined in the above embodiment, a directed edge is first added to the entity node, that is, a connection line with an arrow to represent the data flow direction is added. Figure 2 As shown in the figure, after the image node L0 is processed by radiometric calibration, the image node L1 is obtained, so a directed edge is generated from the image node L0 to the image node L1. After the image node L1 is processed by atmospheric correction, geometric correction and orthorectification, the image node L2 is obtained, so a directed edge is generated from the image node L1 to the image node L2.

[0127] It is understandable that a directed edge only indicates that there is a dependency between two entity nodes, that is, there is a data flow direction, but it cannot directly determine what kind of image processing has been performed. Therefore, in order to achieve the tracing of image task information, in an optional embodiment of constructing a blood relationship graph, the corresponding image task information is added to the directed edge.

[0128] Specifically, you can Figure 2 As shown, the specific image processing method is added to the directed edge for display, or it can be hidden and displayed when the directed edge is clicked. Of course, in another optional embodiment, it can also be added to the corresponding entity node, which is not limited in this application.

[0129] In an optional embodiment, the remote sensing image data management method provided in the present application further includes:

[0130] When receiving an image tracking instruction, parsing the image tracking instruction to determine a starting node for tracking;

[0131] Among all entity nodes in the blood relationship graph, locate the target entity node corresponding to the start node;

[0132] The tracking result of the target entity node is sent to the terminal; wherein the tracking result at least includes all reachable paths between the target entity nodes.

[0133] In a specific embodiment, after the kinship graph of the remote sensing image is generated, in an optional embodiment, the graph database Neo4j can be used for storage, so that the complete path from one entity node to another entity node and its related information can be tracked through the graph query language Cypher provided by Neo4j.

[0134] Specifically, the image tracking instruction input by the user is accepted, and the image tracking instruction is parsed to determine the starting node of the tracking, wherein the starting node includes the start node and the end node. Further, among all the entity nodes in the blood relationship graph, the target entity node corresponding to the starting node is located.

[0135] Then, the tracking result of the queried target entity node is sent to the terminal for the user to view visually. It is understandable that there may be multiple reachable paths between any two nodes, so the tracking result at least includes all reachable paths between the target entity nodes. Of course, in some optional embodiments, the tracking result also includes the data ID, image task information, and node attribute information of each entity node on the reachable path.

[0136] For example, suppose you want to query all paths from image node ImgNode_L0_001 to slice node SliceNode_L1_002_001 in Table 1. The script for querying in Cypher language is: MATCH p = ((img:ImageNode{id:'ImgNode_L0_001'})-[*..]->(slice:SliceNode{id:'SliceNode_L1_002_001'})), RETURNp.

[0137] Among them, MATCH is used to specify the matching keyword in the blood relationship graph, that is, the starting node. Among them, (img:ImageNode{id:'ImgNode_L0_001'}): is the starting node. -[*..]->: represents a path of any length from the starting node to the ending node.

[0138] In p = ((img:ImageNode{id:'ImgNode_L0_001'})-[*..]->(slice:SliceNode{id:'SliceNode_L1_002_001'})), p is a variable used to store the matched reachable path. (slice:SliceNode{id:'SliceNode_L1_002_001'}) is the end node. RETURN p is the returned query result p, which stores all reachable paths from ImgNode_L0_001 to SliceNode_L1_002_001.

[0139] Therefore, the remote sensing image data management method provided in the embodiment of the present application, based on the constructed remote sensing image blood relationship diagram, can realize the rapid and accurate tracking of any image picture, and can timely view and understand key information such as the source, destination, change process, processing flow and processing method of the image picture, so as to provide more accurate and reliable data support for business decisions.

[0140] As remote sensing image data continues to increase, in order to continuously improve the reliability of remote sensing image management and improve the accuracy of remote sensing image data, in an optional embodiment, the remote sensing image data management method provided by the present application further includes:

[0141] When an update instruction to update the current blood relationship graph is obtained; determining whether there is a target image picture that is the same as the newly added image picture in the current blood relationship graph;

[0142] If it exists, determine whether the quality of the newly added image is higher than that of the target image; if it is higher, replace the newly added image with the target image; if it is not higher, discard the newly added image;

[0143] If it does not exist, add the new image to the current blood relationship diagram.

[0144] In a specific embodiment, the blood relationship diagram can be updated in real time, can be updated at regular intervals, or can be updated when an update instruction is received, which is not limited in this application.

[0145] In an optional embodiment, when an update instruction for updating the current blood relationship graph is obtained and the update instruction is a new addition instruction, the update instruction is parsed to determine whether there is a target image picture that is the same as the new addition image picture in the current blood relationship graph.

[0146] If it exists, it is necessary to determine whether the quality of the existing target image is higher than that of the newly added image. Specifically, it can be determined by judging parameters such as resolution. If it is not higher, the original target image is retained. If it is higher, the original target image is replaced by the newly added image, thereby improving the quality of the current blood relationship diagram. It should be noted that when replacing with a newly added image, in addition to replacing the image, the image task information and node attribute information corresponding to the newly added image also need to be replaced.

[0147] If there is no target image identical to the newly added image in the current blood relationship graph, the newly added image will be added to the current blood relationship graph, and the corresponding image task information and node attribute information will be added at the same time.

[0148] It is worth noting that in addition to adding new instructions, update instructions also include but are not limited to modifying instructions and deleting instructions. In an optional embodiment, users can regularly verify and supplement the blood relationship map stored in the graph database Neo4j. Ensure the integrity, reliability and traceability of the data by tracing the data source, data flow path, data dependency and data change records. If data is found to be missing or inconsistent, the blood relationship map must be supplemented, modified and deleted in a timely manner to ensure the integrity and timeliness of the data.

[0149] In the above-mentioned embodiments, the remote sensing image data management method is described in detail. The present application also provides a corresponding embodiment of a remote sensing image data management device.

[0150] Figure 4 A schematic diagram of the structure of a remote sensing image data management device provided in an embodiment of the present application is shown in FIG. Figure 4 As shown, the device comprises:

[0151] The image metadata acquisition module 40 is used to acquire image metadata;

[0152] The entity node definition module 41 is used to define each image picture in the image metadata as an entity node in the blood relationship graph; wherein the entity node includes an image node and a slice node;

[0153] An ID allocation module 42 is used to allocate a unique data ID to a physical node to serve as the DNA of the image;

[0154] A data flow direction determination module 43, used to determine the data flow direction between entity nodes;

[0155] The construction module 44 is used to construct a blood relationship diagram according to the data ID and the data flow direction.

[0156] In addition, the remote sensing image data management device provided in the embodiment of the present application also includes:

[0157] A blood relationship determination module, used to determine a first blood relationship between image nodes and a second blood relationship between image nodes and slice nodes;

[0158] An image ID allocation module, used to allocate a unique image ID to each image node according to the first blood relationship;

[0159] The slice ID allocation module is used to allocate a unique slice ID to each slice node according to the second blood relationship and the image ID; wherein the slice ID includes a part of the image ID.

[0160] The blood relationship determination module is also used to determine the third blood relationship between the parent slice node and the child slice node;

[0161] A parent slice ID allocation module, used to allocate a unique parent slice ID to the parent slice node according to the second blood relationship and the image ID; wherein the parent slice ID includes a part of the image ID;

[0162] The sub-slice ID allocation module is used to allocate a unique sub-slice ID to the sub-slice node according to the third blood relationship; wherein the sub-slice ID includes a part of the parent slice ID.

[0163] An image processing task identification module is used to identify image processing tasks in image metadata;

[0164] A dependency determination module, used to determine whether the generation of one node between any two entity nodes depends on another node according to the image processing task, so as to determine the dependency relationship between any two entity nodes;

[0165] The data flow direction determination module is used to determine the data flow direction according to the dependency relationship.

[0166] An extraction module is used to extract image task information and node attribute information of entity nodes from image metadata; wherein the image task information at least includes image processing timestamp, image processing method, and image production information; the node attribute information at least includes node ownership information, node storage path, and node application type;

[0167] The injection module is used to inject node attribute information into the corresponding entity node;

[0168] The directed edge adding module is used to add directed edges between entity nodes according to the direction of data flow;

[0169] The image task information adding module is used to add corresponding image task information in the directed edges to construct a blood relationship graph.

[0170] A starting node determination module is used to parse the image tracking instruction when receiving the image tracking instruction to determine the starting node of the tracking;

[0171] A positioning module is used to locate the target entity node corresponding to the starting node among all entity nodes in the blood relationship graph;

[0172] The sending module is used to send the tracking result of the target entity node to the terminal; wherein the tracking result at least includes all reachable paths between the target entity nodes.

[0173] The processing module is used to, when an update instruction to update the current blood relationship diagram is obtained, determine whether there is a target image picture that is the same as the newly added image picture in the current blood relationship diagram; if so, determine whether the quality of the newly added image picture is higher than the target image picture; if so, replace the newly added image picture with the target image picture; if not, discard the newly added image picture; if not, add the newly added image picture to the current blood relationship diagram.

[0174] Figure 5 A schematic diagram of a remote sensing image data management device provided in another embodiment of the present application is shown in FIG. Figure 5 As shown, the remote sensing image data management device includes: a memory 50 for storing a computer program;

[0175] The processor 51 is used to implement the steps of the remote sensing image data management method mentioned in the above embodiment when executing the computer program.

[0176] The remote sensing image data management device provided in this embodiment may include but is not limited to a laptop computer or a desktop computer.

[0177] Among them, the processor 51 may include one or more processing cores, such as a 4-core processor, an 8-core processor, etc. The processor 51 can be implemented in at least one hardware form of a digital signal processor (Digital Signal Processor, referred to as DSP), a field programmable gate array (Field-Programmable Gate Array, referred to as FPGA), and a programmable logic array (Programmable Logic Array, referred to as PLA). The processor 51 may also include a main processor and a coprocessor. The main processor is a processor for processing data in the awake state, also known as a central processing unit (Central Processing Unit, referred to as CPU); the coprocessor is a low-power processor for processing data in the standby state. In some embodiments, the processor 51 may be integrated with a graphics processing unit (Graphics Processing Unit, referred to as GPU), and the GPU is responsible for rendering and drawing the content to be displayed on the display screen. In some embodiments, the processor 51 may also include an artificial intelligence (Artificial Intelligence, referred to as AI) processor, which is used to process computing operations related to machine learning.

[0178] The memory 50 may include one or more computer-readable storage media, which may be non-transitory. The memory 50 may also include a high-speed random access memory, and a non-volatile memory, such as one or more disk storage devices, flash memory storage devices. In this embodiment, the memory 50 is at least used to store the following computer program 501, wherein, after the computer program is loaded and executed by the processor 51, it can implement the relevant steps of the remote sensing image data management method disclosed in any of the aforementioned embodiments. In addition, the resources stored in the memory 50 may also include an operating system 502 and data 503, etc., and the storage method may be temporary storage or permanent storage. Among them, the operating system 502 may include Windows, Unix, Linux, etc. Data 503 may include, but is not limited to, relevant data involved in the remote sensing image data management method, etc.

[0179] In some embodiments, the remote sensing image data management device may further include a display screen 52 , an input / output interface 53 , a communication interface 54 , a power supply 55 , and a communication bus 56 .

[0180] Those skilled in the art will understand that Figure 5 The structure shown in the figure does not constitute a limitation on the remote sensing image data management device, and may include more or fewer components than those shown in the figure.

[0181] The remote sensing image data management device provided in the embodiment of the present application includes a memory and a processor. When the processor executes the program stored in the memory, it can implement the remote sensing image data management method in the above embodiment.

[0182] It should be noted that, although the operations are depicted in a specific order in the accompanying drawings, this should not be understood as requiring these operations to be performed in the specific order shown or to be performed sequentially, or requiring all illustrated operations to be performed to achieve the desired results. In some cases, multitasking and parallel processing may be advantageous. In addition, the separation of various system modules and components in the above-described embodiments should not be understood as requiring such separation in all embodiments, and it should be understood that the described program components and systems can generally be integrated together in a single software product, or packaged into multiple software products.

Claims

1. A remote sensing image data management method, characterized in that: The method comprises: Get image metadata; Define each image picture in the image metadata as an entity node in the blood relationship graph; wherein the entity node includes an image node and a slice node; Assigning a unique data ID to the physical node to serve as the DNA of the image; Determining the data flow direction between the entity nodes; The blood relationship graph is constructed according to the data ID and the data flow direction.

2. The remote sensing image data management method according to claim 1, characterized in that: The data ID includes an image ID and a slice ID; the assigning a unique data ID to the entity node includes: Determine a first blood relationship between the image nodes and a second blood relationship between the image nodes and the slice nodes; According to the first blood relationship, assigning a unique image ID to each image node; According to the second blood relationship and the image ID, a unique slice ID is allocated to each slice node; wherein the slice ID includes a part of the image ID.

3. The remote sensing image data management method according to claim 2, characterized in that: If the slice node includes a parent slice node and a child slice node, and the slice ID includes a parent slice ID and a child slice ID; then, according to the second blood relationship and the image ID, a unique slice ID is allocated to each slice node, including: Determine a third blood relationship between the parent slice node and the child slice node; According to the second blood relationship and the image ID, a unique parent slice ID is allocated to the parent slice node; wherein the parent slice ID includes a part of the image ID; According to the third blood relationship, a unique sub-slice ID is assigned to the sub-slice node; wherein the sub-slice ID includes a part of the parent slice ID.

4. The remote sensing image data management method according to claim 1, characterized in that: The determining the data flow direction between the entity nodes includes: identifying an image processing task in the image metadata; According to the image processing task, determining whether the generation of one node between any two of the entity nodes depends on the other node, so as to determine the dependency relationship between any two of the entity nodes; The data flow direction is determined according to the dependency relationship.

5. The remote sensing image data management method according to claim 1, characterized in that: The step of constructing the blood relationship graph according to the data ID and the data flow direction includes: Extracting image task information and node attribute information of the entity node from the image metadata; wherein the image task information at least includes image processing timestamp, image processing method, and image production information; and the node attribute information at least includes node ownership information, node storage path, and node application type; Injecting the node attribute information into the corresponding entity node; According to the data flow direction, adding directed edges between the entity nodes; The corresponding image task information is added to the directed edges to construct the blood relationship graph.

6. The remote sensing image data management method according to claim 1, characterized in that: The method further comprises: When receiving an image tracking instruction, parsing the image tracking instruction to determine a starting node for tracking; Among all entity nodes in the blood relationship graph, locate the target entity node corresponding to the start node; The tracking result of the target entity node is sent to a terminal; wherein the tracking result at least includes all reachable paths between the target entity nodes.

7. The remote sensing image data management method according to claim 1, characterized in that: The method further comprises: When an update instruction to update the current blood relationship graph is obtained; Determine whether there is a target image picture that is the same as the newly added image picture in the current blood relationship graph; If so, determine whether the quality of the newly added image is higher than that of the target image; if so, replace the newly added image with the target image; if not, discard the newly added image; If not present, the newly added image picture is added to the current blood relationship diagram.

8. A remote sensing image data management device, characterized in that: The device comprises: An image metadata acquisition module is used to acquire image metadata; An entity node definition module, used to define each image picture in the image metadata as an entity node in the blood relationship graph; wherein the entity node includes an image node and a slice node; An ID allocation module, used to allocate a unique data ID to the entity node to serve as the DNA of the image; A data flow direction determination module, used to determine the data flow direction between the entity nodes; A construction module is used to construct the blood relationship diagram according to the data ID and the data flow direction.

9. A remote sensing image data management device, comprising a memory and a processor, wherein the memory stores a computer program that can be run on the processor, characterized in that: When the processor executes the program, the steps of the remote sensing image data management method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the remote sensing image data management method according to any one of claims 1 to 7 are implemented.

Images