Remote sensing target knowledge graph dynamic error correction method based on multi-source sensor fusion

Abstract

This invention provides a dynamic error correction method for remote sensing target knowledge graphs based on multi-source sensor fusion. It acquires attribute information and category information of remote sensing targets detected by multiple sensors; fuses the attribute information detected by each sensor to obtain fused attribute information; inputs the fused attribute information into the remote sensing target knowledge graph to obtain a first classification result; determines the error type of the remote sensing target knowledge graph based on the first classification result and category information; and performs error correction processing on the remote sensing target knowledge graph according to the error type and preset error correction rules. This method enables the identification of remote sensing targets using data from multiple remote sensing sources, filling a gap in the current field of error correction for remote sensing target knowledge graphs. By using multi-source remote sensing data as input data for the error correction algorithm, it fully integrates the different characteristics of target objects exhibited by different sensors, records target attribute information in a richer dimension, and improves the target recognition rate.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, specifically to a dynamic error correction method for remote sensing target knowledge graphs based on multi-source sensor fusion. Background Technology

[0002] Knowledge graphs are structured semantic knowledge bases used to rapidly describe concepts and their relationships in the physical world. Knowledge graphs can aggregate vast amounts of information, data, and links into knowledge, making information resources easier to compute, understand, and evaluate, and enabling rapid knowledge response and reasoning. Remote sensing target identification has long faced challenges due to its rich background information, large data volume, and difficulty in rapid identification. Establishing a remote sensing knowledge base to support spatiotemporal big data mining is a convenient method. Therefore, research on the construction of remote sensing target knowledge graphs has emerged. While there are existing research results on the construction of remote sensing information knowledge graphs, even well-constructed knowledge graphs inevitably have some shortcomings.

[0003] In the actual process of building knowledge graphs, it has been found that the quality of real-world data is often low, which seriously affects the quality of data analysis and mining results, leading to unreliable outcomes. Data collection and acquisition often introduce data errors, such as missing values, incorrect input, inconsistent formatting, duplicate data for the same real-world entity, and data that violates business and data integrity rules. The meaning represented by this erroneous data does not match the real world, so the final decision is easily influenced by this erroneous data, leading to decision-making errors or biases. Currently, there is no good solution.

[0004] Currently, knowledge graph error correction algorithms are not applicable to the remote sensing data field. Due to the inherent characteristics of remote sensing data, existing algorithms cannot be directly and easily ported. Existing knowledge graph completion algorithms only address the problem of completing partially missing triples and do not offer effective algorithms for supplementing and correcting newly emerging knowledge in the knowledge graph. Summary of the Invention

[0005] To address the aforementioned problems in existing technologies, this invention provides a dynamic error correction method for remote sensing target knowledge graphs based on multi-source sensor fusion. Specifically, it includes:

[0006] In a first aspect, the present invention provides a dynamic error correction method for remote sensing target knowledge graphs based on multi-source sensor fusion, comprising:

[0007] Acquire attribute information and category information of remote sensing targets detected by multiple sensors;

[0008] By fusing the attribute information detected by each sensor, the fused attribute information of the remote sensing target is obtained;

[0009] The fused attribute information of the remote sensing targets is input into the remote sensing target knowledge graph to obtain the first classification result;

[0010] Based on the first classification result and the category information of the remote sensing target, determine the error type of the remote sensing target knowledge graph;

[0011] The remote sensing target knowledge graph is corrected according to the error types and preset error correction rules.

[0012] Secondly, the present invention also provides a dynamic error correction device for a remote sensing target knowledge graph based on multi-source sensor fusion, comprising:

[0013] The acquisition module is used to acquire attribute information and category information of remote sensing targets detected by multiple sensors;

[0014] The processing module is used to fuse the attribute information detected by each sensor to obtain the fused attribute information of the remote sensing target;

[0015] The processing module is also used to input the fused attribute information of remote sensing targets into the remote sensing target knowledge graph to obtain the first classification result;

[0016] The processing module is also used to determine the error type of the remote sensing target knowledge graph based on the first classification result and the category information of the remote sensing target;

[0017] The processing module is also used to perform error correction processing on the remote sensing target knowledge graph according to the error type and preset error correction rules.

[0018] Thirdly, the present invention also provides an electronic device, including a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus;

[0019] Memory, used to store computer programs;

[0020] The processor, when executing a program stored in memory, implements any of the methods provided in the first aspect.

[0021] Fourthly, the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements any of the methods provided in the first aspect.

[0022] The beneficial effects of this invention are:

[0023] This invention provides a dynamic error correction method for remote sensing target knowledge graphs based on multi-source sensor fusion. This method acquires attribute information and category information of remote sensing targets detected by multiple sensors; fuses the attribute information detected by each sensor to obtain fused attribute information of the remote sensing targets; inputs the fused attribute information of the remote sensing targets into a remote sensing target knowledge graph to obtain a first classification result; determines the error type of the remote sensing target knowledge graph based on the first classification result and the category information of the remote sensing targets; and performs error correction processing on the remote sensing target knowledge graph according to the error type and preset error correction rules. This method can identify remote sensing targets using data from multiple remote sensing sources, filling a gap in the current field of remote sensing target knowledge graph error correction. By using multi-source remote sensing data as input data for the error correction algorithm, it fully integrates the different characteristics of target objects exhibited by different sensors, records the attribute information of each target in a richer dimension, and improves the target recognition rate.

[0024] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0025] Figure 1 A flowchart illustrating a dynamic error correction method for remote sensing target knowledge graphs based on multi-source sensor fusion provided by this invention;

[0026] Figure 2 This invention provides a schematic diagram of a remote sensing target knowledge graph.

[0027] Figure 3 A flowchart illustrating another method for dynamic error correction of remote sensing target knowledge graphs based on multi-source sensor fusion provided by the present invention;

[0028] Figure 4 A flowchart illustrating another method for dynamic error correction of remote sensing target knowledge graphs based on multi-source sensor fusion provided by the present invention;

[0029] Figure 5 This is a schematic diagram of the structure of a dynamic error correction device for remote sensing target knowledge graph based on multi-source sensor fusion, provided by the present invention. Detailed Implementation

[0030] The present invention will be further described in detail below with reference to specific embodiments, but the implementation of the present invention is not limited thereto.

[0031] In text-based knowledge graph error correction, researchers input the same data into the same knowledge graph multiple times to obtain the output results each time. If the results are inconsistent, the knowledge graph is considered to be erroneous, and the most frequently occurring result is deemed correct. Currently, most existing solutions for knowledge graph correction focus on knowledge graph completion. Knowledge graphs contain numerous fact triples (head entity, relation, tail entity). When triples are missing, knowledge graph completion algorithms can be used. For example, knowledge graph embedding models embed entities and relations as low-dimensional representations into the knowledge graph and define a scoring function in the embedding space to measure the reasonableness of the triples. Generally, valid triples should score higher than invalid triples. However, knowledge completion algorithms cannot be used to correct knowledge graphs when faced with completely missing triples.

[0032] To address the problems existing in the prior art, this invention provides a dynamic error correction method for remote sensing target knowledge graphs based on multi-source sensor fusion. This method first identifies the error types of the remote sensing target knowledge graph, and then performs correction processing according to the error types.

[0033] like Figure 1 As shown, the method includes:

[0034] S101. Acquire the attribute information and category information of remote sensing targets detected by multiple sensors.

[0035] Optionally, the attribute information of the remote sensing target may include one or more of the following: shape outline, color, motion, spectrum, spatial spectrum, and dimensional information.

[0036] Sensors include, but are not limited to, optical sensors, radar sensors, infrared sensors, microwave radiometers, and lidar. Optical sensors include, but are not limited to, multispectral, hyperspectral, and hyperspectral sensors. The category information of remotely sensed targets represents their true categories.

[0037] S102. Fuse the attribute information detected by each sensor to obtain the fused attribute information of the remote sensing target.

[0038] The attribute information of remote sensing targets obtained from different sensors is fused, and the fused attribute information is used as input data for the remote sensing target knowledge graph.

[0039] S103. Input the fused attribute information of the remote sensing target into the remote sensing target knowledge graph to obtain the first classification result.

[0040] For example, knowledge graphs such as Figure 2As shown, specific objects such as "forest" and "water body" are entities in the remote sensing target knowledge graph, while "shape" and "color" represent the relationships between nodes. The continuous sequence from one entity to another, "optical sensor -> detection -> forest -> color -> green -> broadleaf forest," is a path in the graph, indicating that a forest was detected by an optical sensor, its color and other characteristics were analyzed, and tree species were classified.

[0041] S104. Based on the first classification result and the category information of the remote sensing target, determine the error type of the remote sensing target knowledge graph.

[0042] S105. Perform error correction processing on the remote sensing target knowledge graph according to the error types and preset error correction rules.

[0043] For example, the error types of remote sensing target knowledge graphs include Type I errors and Type II errors. Type I errors occur when the identification results obtained from different sensor signals are not completely the same, requiring modification of the existing knowledge graph and the addition of new path information. Type II errors occur when the identified target belongs to the missing information in the knowledge graph. In this case, the knowledge graph lacks the attribute or category information of this target, requiring the addition of new content to the knowledge graph.

[0044] Accordingly, regarding S104 of the above embodiment, which determines the error type of the remote sensing target knowledge graph based on the first classification result and the category information of the remote sensing target, optionally, see [link to relevant documentation]. Figure 3 Based on the first classification result and the category information of remote sensing targets, the error types of the remote sensing target knowledge graph are determined, including:

[0045] S301. Compare the first classification result with the category information of the remote sensing target. If the first classification is inconsistent with the category information, input the attribute information detected by each sensor into the remote sensing target knowledge graph to obtain the second classification result corresponding to each attribute information.

[0046] Specifically, if the first classification is inconsistent with the category information, it is determined that there is an error in the remote sensing target knowledge graph, which needs to be corrected; if the first classification is consistent with the category information, it is determined that there is no error in the remote sensing target knowledge graph, which does not need to be corrected.

[0047] S302. Determine whether the results of each second classification are not completely the same. If so, proceed to S303; otherwise, proceed to S304.

[0048] S303. The error type of the remote sensing target knowledge graph is determined to be a Type I error.

[0049] S304. The error type of the remote sensing target knowledge graph is determined to be a Type II error.

[0050] Furthermore, regarding S105 in the above embodiment, which performs error correction processing on the remote sensing target knowledge graph based on the error type of the remote sensing target knowledge graph and the preset error correction rules, in one possible implementation, when the error type of the remote sensing target knowledge graph is a first-type error, the error correction processing on the remote sensing target knowledge graph based on the error type of the remote sensing target knowledge graph and the preset error correction rules includes: based on the knowledge graph completion algorithm, completing the attribute information of the triple corresponding to the remote sensing target in the remote sensing target knowledge graph.

[0051] Optionally, the attribute information of the corresponding triples of the remote sensing target in the remote sensing target knowledge graph can be supplemented, including modifying the path information in the remote sensing target knowledge graph, or adding new path information in the remote sensing target knowledge graph.

[0052] Specifically, knowledge graph completion algorithms, such as the TranE algorithm, are used. These algorithms transform the original entities (targets, attributes, etc.) and relations ("color", "shape", etc.) of the knowledge graph into low-dimensional vectors, treating relations as translation vectors from the head entity to the tail entity. For example, in "forest + color = green", the "color" relation vector can explain the difference between the "forest" and "green" vectors. The algorithm optimizes the objective function to ensure that the head entity vector, relation vector, and tail entity vector are as close as possible to a closed loop. Therefore, if any element in a triplet of the remote sensing target knowledge graph is missing, the algorithm will find the element most likely to complete the vector closure; or if there is an error in the triplet, the algorithm will select the correct entity that is closer to the vector, thus updating the knowledge graph.

[0053] Furthermore, regarding S105 in the above embodiment, which performs error correction processing on the remote sensing target knowledge graph based on the error type and preset error correction rules, in one possible implementation, see [link to relevant documentation]. Figure 4 When the error type of the remote sensing target knowledge graph is Type II error, the remote sensing target knowledge graph is corrected according to the error type and preset correction rules, including:

[0054] S401. Based on the results of each second classification, determine whether there is erroneous path information in the remote sensing target knowledge graph. If so, proceed to S402; otherwise, proceed to S403.

[0055] S402. Delete incorrect path information.

[0056] S403. Based on reinforcement learning, determine the missing information of the remote sensing target knowledge graph, and determine new triples related to the remote sensing target based on the missing information.

[0057] S404. Add the new triplet to the remote sensing target knowledge graph.

[0058] Optionally, based on reinforcement learning, the missing information of the remote sensing target knowledge graph is determined, and new triples related to the remote sensing target are determined based on the missing information. This includes: constructing a reinforcement learning model based on the DeepPath algorithm according to the results of each second classification, determining the missing information of the remote sensing target knowledge graph through the reinforcement learning model, and determining new triples related to the remote sensing target based on the missing information.

[0059] Specifically, the current knowledge graph and identified missing category information are used as input data. The DeepPath algorithm is used for path exploration. This algorithm models the path search and planning problem in the knowledge graph as a reinforcement learning problem, and the policy network acts as an agent to reason within the knowledge graph. The policy network outputs the probability of taking different paths based on nodes and their existing relationships, and selects the next relationship for the current node based on the probability. During this process, if the path accurately reflects the categories of the old and newly added data, a reward is given, and this is fed back to the policy network to adjust its parameters. The maximum cumulative reward is obtained when the knowledge graph update is completed.

[0060] Since path-based reasoning in remote sensing target knowledge graphs requires finding a multi-hop path from the head entity to the tail entity within the knowledge graph, this path establishment process can be viewed as a Markov decision process. Therefore, reinforcement learning can be used to model this Markov decision process. Using reinforcement learning in remote sensing target knowledge graph error correction can lead to the creation of a knowledge graph with more concise paths.

[0061] This invention provides a dynamic error correction method for remote sensing target knowledge graphs based on multi-source sensor fusion. This method acquires attribute information and category information of remote sensing targets detected by multiple sensors; fuses the attribute information detected by each sensor to obtain fused attribute information of the remote sensing targets; inputs the fused attribute information of the remote sensing targets into a remote sensing target knowledge graph to obtain a first classification result; determines the error type of the remote sensing target knowledge graph based on the first classification result and the category information of the remote sensing targets; and performs error correction processing on the remote sensing target knowledge graph according to the error type and preset error correction rules. This method can identify remote sensing targets using data from multiple remote sensing sources, filling a gap in the current field of remote sensing target knowledge graph error correction. By using multi-source remote sensing data as input data for the error correction algorithm, it fully integrates the different characteristics of target objects exhibited by different sensors, records the attribute information of each target in a richer dimension, and improves the target recognition rate.

[0062] Figure 5 This is a schematic diagram of the structure of a dynamic error correction device for remote sensing target knowledge graphs based on multi-source sensor fusion, as provided by the present invention. Figure 5As shown, the device includes:

[0063] The acquisition module 51 is used to acquire attribute information and category information of remote sensing targets detected by multiple sensors.

[0064] The processing module 52 is used to fuse the attribute information detected by each sensor to obtain the fused attribute information of the remote sensing target.

[0065] The processing module 52 is also used to input the fused attribute information of the remote sensing target into the remote sensing target knowledge graph to obtain the first classification result.

[0066] The processing module 52 is also used to determine the error type of the remote sensing target knowledge graph based on the first classification result and the category information of the remote sensing target.

[0067] The processing module 52 is also used to perform error correction processing on the remote sensing target knowledge graph according to the error type and preset error correction rules.

[0068] This invention also provides an electronic device structure, including a processor, a communication interface, a memory, and a communication bus, wherein the processor, communication interface, and memory communicate with each other via the communication bus.

[0069] Memory, used to store computer programs;

[0070] When a processor executes a program stored in memory, it implements the steps provided in the above method embodiments.

[0071] The communication interface is used for communication between the aforementioned electronic devices and other devices.

[0072] The method provided in this invention can be applied to electronic devices. Specifically, the electronic device can be a desktop computer, a portable computer, a smart mobile terminal, a server, etc. No limitation is made herein; any electronic device that can implement this invention falls within the protection scope of this invention.

[0073] The present invention also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps provided in the above-described method embodiments.

[0074] For the embodiments of the device / electronic device / storage medium, since they are basically similar to the method embodiments, the description is relatively simple. For specific details and beneficial effects, please refer to the description of the method embodiments.

[0075] The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0076] The above description, in conjunction with specific preferred embodiments, provides a further detailed explanation of the present invention. It should not be construed that the specific implementation of the present invention is limited to these descriptions. For those skilled in the art, various simple deductions or substitutions can be made without departing from the concept of the present invention, and all such modifications and substitutions should be considered within the scope of protection of the present invention.

Claims

1. A dynamic error correction method for remote sensing target knowledge graphs based on multi-source sensor fusion, characterized in that, include: Acquire attribute information and category information of remote sensing targets detected by multiple sensors; By fusing the attribute information detected by each of the sensors, the fused attribute information of the remote sensing target is obtained; The fused attribute information of the remote sensing target is input into the remote sensing target knowledge graph to obtain the first classification result; Based on the first classification result and the category information of the remote sensing target, the error type of the remote sensing target knowledge graph is determined, including: Compare the first classification result with the category information of the remote sensing target. If the first classification is inconsistent with the category information, then input the attribute information detected by each sensor into the remote sensing target knowledge graph to obtain the second classification result corresponding to each attribute information. Determine whether the results of each of the second classifications are not completely the same. If so, determine that the error type of the remote sensing target knowledge graph is a first type of error. Otherwise, determine that the error type of the remote sensing target knowledge graph is a second type of error. The remote sensing target knowledge graph is corrected according to the error type and preset error correction rules. When the error type of the remote sensing target knowledge graph is Type I error The step of performing error correction processing on the remote sensing target knowledge graph based on the error type and preset error correction rules includes: Based on the knowledge graph completion algorithm, the attribute information of the triples corresponding to the remote sensing target in the remote sensing target knowledge graph is completed; When the error type of the remote sensing target knowledge graph is Type II error The step of performing error correction processing on the remote sensing target knowledge graph based on the error type and preset error correction rules includes: S41, Based on each of the second classification results, determine whether there is erroneous path information in the remote sensing target knowledge graph. If so, proceed to S42; otherwise, proceed to S43. S42, Delete the erroneous path information; S43, Based on reinforcement learning, determine the missing information of the remote sensing target knowledge graph, and determine new triples related to the remote sensing target based on the missing information; S44, the new triple is added to the remote sensing target knowledge graph.

2. The method according to claim 1, characterized in that, The step of determining missing information in the remote sensing target knowledge graph based on reinforcement learning, and determining new triples related to the remote sensing target based on the missing information, includes: Based on the DeepPath algorithm, a reinforcement learning model is constructed according to each of the second classification results. The missing information of the remote sensing target knowledge graph is determined through the reinforcement learning model, and new triples related to the remote sensing target are determined based on the missing information.

3. The method according to claim 2, characterized in that, The attribute information of the remote sensing target includes one or more of the following: shape outline, color, motion, spectrum, spatial spectrum, and dimensional information.

4. A dynamic error correction device for remote sensing target knowledge graphs based on multi-source sensor fusion, characterized in that, include: The acquisition module is used to acquire attribute information and category information of remote sensing targets detected by multiple sensors; The processing module is used to fuse the attribute information detected by each of the sensors to obtain the fused attribute information of the remote sensing target; The processing module is also used to input the fused attribute information of the remote sensing target into the remote sensing target knowledge graph to obtain the first classification result; The processing module is further configured to determine the error type of the remote sensing target knowledge graph based on the first classification result and the category information of the remote sensing target, including: Compare the first classification result with the category information of the remote sensing target. If the first classification is inconsistent with the category information, then input the attribute information detected by each sensor into the remote sensing target knowledge graph to obtain the second classification result corresponding to each attribute information. Determine whether the results of each of the second classifications are not completely the same. If so, determine that the error type of the remote sensing target knowledge graph is a first type of error. Otherwise, determine that the error type of the remote sensing target knowledge graph is a second type of error. The processing module is also used to perform error correction processing on the remote sensing target knowledge graph according to the error type and preset error correction rules; When the error type of the remote sensing target knowledge graph is Type I error The step of performing error correction processing on the remote sensing target knowledge graph based on the error type and preset error correction rules includes: Based on the knowledge graph completion algorithm, the attribute information of the triples corresponding to the remote sensing target in the remote sensing target knowledge graph is completed; When the error type of the remote sensing target knowledge graph is Type II error The step of performing error correction processing on the remote sensing target knowledge graph based on the error type and preset error correction rules includes: S41, Based on each of the second classification results, determine whether there is erroneous path information in the remote sensing target knowledge graph. If so, proceed to S42; otherwise, proceed to S43. S42, Delete the erroneous path information; S43, Based on reinforcement learning, determine the missing information of the remote sensing target knowledge graph, and determine new triples related to the remote sensing target based on the missing information; S44, the new triple is added to the remote sensing target knowledge graph.

5. An electronic device, characterized in that, It includes a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; Memory, used to store computer programs; A processor, when executing a program stored in memory, implements the method of any one of claims 1-3.

6. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the method described in any one of claims 1-3.

Images