Containerized space-based heterogeneous multi-source information association fusion method and device

Abstract

The invention relates to a containerized space-based heterogeneous multi-source information association fusion method and device, and the method comprises the steps: firstly completing the access of track and light / SAR decision-level information, then carrying out the space-time alignment of a track, carrying out the track-track association matching and fusion complementation, and completing the batch grouping and warehousing of the track; correlation degree calculation is carried out on the light / SAR slices, and then light / SAR fusion attribute confirmation and target cataloguing storage are completed; and finally, updating and maintaining the situation map to complete generation and distribution of the situation package. According to the device, containerization deployment implementation can be carried out based on edge microclouds, and target position and type estimation can be obtained more accurately based on space-based heterogeneous multi-source information through information association fusion.

Description

Technical Field

[0001] This invention relates to the field of information fusion technology, and in particular to a containerized space-based heterogeneous multi-source information association and fusion method and apparatus. Background Technology

[0002] With the development of aerospace technology, space-based remote sensing systems have become a core platform for information acquisition and processing. They provide highly timely and wide-coverage Earth environment monitoring capabilities through multi-source heterogeneous data (including optical remote sensing images and synthetic aperture radar (SAR) data), and are widely used in disaster early warning, military reconnaissance, climate change analysis, and intelligent traffic management. However, the core challenge facing space-based information processing lies in the efficient correlation and fusion of heterogeneous multi-source data, which directly determines the real-time performance, accuracy, and system robustness of information utilization. Currently, mainstream space-based information processing technologies mainly rely on ground-based centralized processing architectures. Specifically, the satellite platform transmits raw data to ground stations via downlink, where it undergoes manual preprocessing, format conversion, and algorithm fusion to generate the final product. This model has shortcomings in terms of timeliness, heterogeneous compatibility, and resource efficiency, making it difficult to support the evolutionary needs of future space-based intelligent systems. Although edge computing and containerization technologies have made progress in the terrestrial domain, their application in space-based scenarios remains unexplored. Therefore, an innovative approach is urgently needed to deeply integrate containerization with space-based information processing to achieve lightweight, dynamic, and interconnected fusion of heterogeneous data, thereby overcoming the aforementioned bottlenecks. Summary of the Invention

[0003] To address the technical problems existing in the prior art, the present invention aims to provide a containerized space-based heterogeneous multi-source information association and fusion method, which can solve the shortcomings of the prior art in terms of timeliness, heterogeneous compatibility, and resource efficiency, and provide a new generation solution for aerospace information processing.

[0004] To achieve the above-mentioned objectives, this invention provides a containerized space-based heterogeneous multi-source information correlation and fusion method comprising the following steps:

[0005] Step S1: Analyze the trajectory data of the sea surface target, and store the acquired raw trajectory decision information into the database to complete the information access;

[0006] Step S2: Perform spatiotemporal alignment of the acquired original track decision information;

[0007] Step S3: Perform track-track association matching on the completed spatiotemporal alignment information;

[0008] Step S4: Perform track fusion and completion based on the track matching results;

[0009] Step S5: Batch the completed track fusion and data entry into the database.

[0010] Step S6: Slice and analyze the optical / SAR of the sea surface target to complete the access of optical / SAR decision-level information;

[0011] Step S7: Calculate the correlation between optical / SAR slices and identify target locations;

[0012] Step S8: Based on the spatiotemporal location and attribute matching degree, perform point-track association matching between optical / SAR target location information and flight track;

[0013] Step S9: Confirm the optical / SAR fusion attributes of the matched targets;

[0014] Step S10: Associate the target trajectory with the existing situation map and catalog the target into the database;

[0015] Step S11: Update and maintain the situation map based on the situation information recorded on the ground;

[0016] Step S12: Generate and distribute situation packets based on the situation map.

[0017] According to one technical solution of the present invention, in the track-track association matching in step S3, Mahalanobis distance is used to measure the similarity of the state estimates of short tracks, the Hungarian algorithm is used to solve the many-to-many assignment problem to optimize the association, and a sequential pairwise association method is adopted to perform track association, ultimately achieving global track association.

[0018] According to one technical solution of the present invention, in step S4, the track fusion and completion includes a fusion operation of overlapping time periods and a completion operation of missing time periods.

[0019] The fusion operation for the overlapping time period of the tracks includes weighted fusion of overlapping tracks associated with the same target;

[0020] The missing time period completion operation includes interpolating or predicting the trajectory of the missing time period to complete it.

[0021] According to one technical solution of the present invention, step S5 specifically includes:

[0022] Each independent and complete track is assigned a unique batch number. Each independent and complete track includes the starting track, the bifurcation track, and the disappearing track. New tracks that cannot be associated with existing batch numbers are assigned new batch numbers.

[0023] According to one technical solution of the present invention, in step S7, the optical / SAR slice correlation calculation includes: calculating the correlation degree of the same modal slices or cross-modal multimodal slices that are initially screened according to spatiotemporal location under different viewing angles and different lighting conditions as the same target.

[0024] According to one technical solution of the present invention, step S8 specifically includes:

[0025] Point-track association matching involves associating the target location information identified by optical / SAR detailed investigation with existing tracks based on spatiotemporal location and attribute matching information, thus associating the tracks with target identity information. Spatiotemporal location refers to the proximity of the target and track in time and space; attribute matching degree refers to the degree of matching between the attribute information of the target and track.

[0026] According to one technical solution of the present invention, step S11 specifically includes:

[0027] Based on the ground-based situation, the situation map is updated and maintained incrementally using the ground-based situation as the situation anchor point, and the new flight paths and points are associated with the targets in the original situation map.

[0028] According to one aspect of the present invention, a containerized space-based heterogeneous multi-source information correlation and fusion device is provided for the above-described method, comprising:

[0029] The trajectory association and fusion function container is used to perform spatiotemporal alignment of the accessed decision-level information, real-time trajectory-trajectory association matching and fusion completion, and complete the batch compilation and storage of trajectories.

[0030] The attribute joint confirmation function container is used to perform point-track association matching and optical / SAR fusion attribute confirmation on the accessed optical / SAR information and the track information in the database, and to associate the target after fusion attribute confirmation with the existing situation map and catalog it into the database.

[0031] The situation generation function management container is used to update and maintain situation maps based on ground-based situation data, and to generate and distribute situation packages.

[0032] According to one technical solution of the present invention, the containerized space-based heterogeneous multi-source information association and fusion device is deployed and implemented in a containerized manner based on edge micro-cloud.

[0033] Compared with the prior art, the present invention has the following beneficial effects:

[0034] The containerized space-based heterogeneous multi-source information association and fusion method and apparatus provided by the present invention associates and fuses the trajectory data of sea surface targets, and associates and matches the trajectory information of the targets with the point trace information based on the spatiotemporal position and attribute matching degree of the targets, so as to obtain target position and type estimation more accurately based on space-based heterogeneous multi-source information.

[0035] In this invention, the device is deployed and implemented in a containerized manner based on edge micro-cloud, which can realize a unified operating environment and flexible scheduling operation mode. Through the containerized space-based heterogeneous multi-source information association and fusion architecture, it realizes lightweight and dynamic association and fusion of heterogeneous data, solves the shortcomings of existing technologies in terms of timeliness, heterogeneous compatibility and resource efficiency, and provides a new generation of solutions for aerospace information processing. Attached Figure Description

[0036] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the embodiments will be briefly described below. Obviously, the drawings described below are merely some embodiments of the present invention, and those skilled in the art can obtain other drawings based on these drawings without any creative effort.

[0037] Figure 1 This schematic diagram illustrates a flowchart of a containerized space-based heterogeneous multi-source information correlation and fusion method provided in one embodiment of the present invention;

[0038] Figure 2 This diagram illustrates the workflow of a containerized space-based heterogeneous multi-source information correlation and fusion device provided in one embodiment of the present invention.

[0039] Figure 3 This diagram illustrates a containerized deployment method based on an edge micro-cloud according to an embodiment of the present invention. Detailed Implementation

[0040] The description of the embodiments in this specification should be taken in conjunction with the accompanying drawings, which should form part of the complete specification. In the drawings, the shape or thickness of the embodiments may be exaggerated and may be indicated in a simplified or convenient manner. Furthermore, parts of the various structures in the drawings will be described separately; it is worth noting that elements not shown in the figures or not described in words are in a form known to those skilled in the art.

[0041] The descriptions of the embodiments herein, including any references to directions and orientations, are for ease of description only and should not be construed as limiting the scope of the invention. The following description of preferred embodiments involves combinations of features, which may exist independently or in combination; the invention is not particularly limited to the preferred embodiments. The scope of the invention is defined by the claims.

[0042] like Figure 1 As shown, the containerized space-based heterogeneous multi-source information correlation and fusion method provided by the present invention specifically includes:

[0043] Step S1: Analyze the trajectory data of the sea surface target, and store the acquired raw trajectory decision information into the database to complete the information access;

[0044] Specifically, a scripting language is used to parse sea surface target data, and the acquired raw decision-level information such as track, target type, and attributes is stored in the database according to a predefined database table structure format to complete information access.

[0045] Step S2: Perform spatiotemporal alignment of the acquired original track decision information;

[0046] Specifically, the temporal and spatial system deviations of the tracks are corrected, the time points of the overlapping track data are aligned to a fixed interval, and the latitude and longitude positions of the targets are interpolated to align them in time and space, which facilitates subsequent track association and fusion processing.

[0047] Step S3: Perform track-track association matching on the completed spatiotemporal alignment information;

[0048] In step S3, based on the provided target location, target type, attributes, etc., and according to the probability model, segmented short tracks are matched and associated with long tracks based on the criterion of maximizing probability. The core of track association lies in determining whether two short tracks belong to the same target, which requires comprehensive consideration of time, space, and attribute information. Probabilistic methods are particularly suitable for handling complex scenarios with multiple targets and multiple sources. In the track-track association matching in step S3, Mahalanobis distance is used to measure the similarity of the state estimates of short tracks, the Hungarian algorithm is used to solve the many-to-many assignment problem to optimize association, and a sequential pairwise association method is used for track association, ultimately achieving global track association. Mahalanobis distance is widely used to measure the similarity of state estimates, while the Hungarian algorithm is used to solve the many-to-many assignment problem to optimize association. The sequential pairwise association method is used for track association to effectively improve computational efficiency and real-time performance, ultimately achieving global track association.

[0049] Step S3 specifically includes the following steps:

[0050] Step S31: Divide all short tracks according to time windows, filter out track segments with completely consistent target types, and divide them into a set of tracks to be matched and a set of matched long track ends.

[0051] Step S32: Calculate the similarity of the state estimation of the track segment using Mahalanobis distance;

[0052] For the "starting point of the track segment to be matched" and the "end point of the matched long track", calculate the Mahalanobis distance and perform similarity calculation for the state estimation of the track segment.

[0053] Step S33: Adopt a sequential pairwise association strategy and proceed from early to late according to time: only match the track to be matched in the current time window and the end of the long track that has been spliced ​​in the previous time; remove track pairs that are not continuous in time or have conflicting attributes, set their matching cost to infinity, and finally generate the track matching cost matrix.

[0054] Step S34: Input the cost matrix into the Hungarian algorithm. The algorithm will solve for the pairing scheme with the minimum overall matching cost under the constraint that "only one track to be matched is matched at the end of a track", and obtain the globally optimal track association result.

[0055] Step S35: Patch the successfully matched track segments onto the long track in chronological order.

[0056] Step S4: Perform track fusion and completion based on the track matching results;

[0057] In step S4, track fusion and completion includes fusion operations for overlapping track time periods and completion operations for missing time periods. Specifically, the fusion operation for overlapping track time periods includes weighted fusion of overlapping tracks associated with the same target to improve positioning accuracy; the completion operation for missing time periods includes interpolation or prediction completion of tracks in missing time periods.

[0058] Step S4 specifically includes:

[0059] Step S41: Based on the obtained track correlation, acquire all track data associated with the same target. Track data includes state estimates (such as position, velocity, etc.) and confidence indices (such as covariance matrix or error estimates).

[0060] Step S42: Determine the time series of the flight path and identify overlapping and missing time periods.

[0061] Step S43: For tracks in overlapping time periods, use a weighted geometric mean fusion method to integrate the state estimates of each track in order to improve positioning accuracy.

[0062] Step S44: For the missing time period of the track, use interpolation method to complete the missing state estimate.

[0063] Step S45: Output the fused complete track, including the fusion results of overlapping time periods and the completion results of missing time periods.

[0064] Step S5: Batch the completed track fusion and data entry into the database;

[0065] Specifically, a unique batch number is assigned to each independent and complete track, which includes the initial track, the bifurcation track, and the disappearance track; new batch numbers are assigned to newly created tracks that cannot be associated with existing batch numbers.

[0066] Step S6: Slice and analyze the optical / SAR data of sea surface targets and complete the access to optical / SAR decision-level information;

[0067] Specifically, the optical / SAR of sea surface targets is sliced, and the optical / SAR point information of sea surface targets is parsed. The original decision-level information such as target traces, target types, and attributes acquired by the optical / SAR payload is stored in the database in a specified format to complete the access of optical / SAR decision-level information.

[0068] Step S7: Perform optical / SAR slice correlation calculation and optical / SAR detailed survey target location identification;

[0069] In step S7, the optical / SAR slice correlation calculation includes: calculating the correlation between optical-optical, SAR-SAR, and optical-SAR slices of the same or cross-modal multimodal nature that have passed the initial spatiotemporal location screening and are of the same target, in order to improve the accuracy of point / track correlation and improve the accuracy of target identity category recognition.

[0070] Step S7 specifically includes:

[0071] Step S71: Based on the spatiotemporal location, filter out the same-modal slices and cross-modal multimodal slices that may be the same target from different perspectives and under different lighting conditions;

[0072] Step S72: Calculate the correlation degree of the same target for the selected same-modal slices and cross-modal multimodal slices. When the correlation degree calculation result exceeds the correlation degree threshold, they are regarded as the same target, thereby improving the accuracy of point / track correlation and improving the accuracy of target identity category recognition.

[0073] Step S73: Perform optical / SAR detailed survey target location identification on the associated optical / SAR slices to obtain detailed survey target location information.

[0074] Step S8: Based on the spatiotemporal location and attribute matching degree, perform point-track association matching between the target location information and the track in the optical / SAR detailed investigation;

[0075] In step S8, the point-track association matching is to associate the target point information identified by light / SAR detailed investigation with the existing track based on spatiotemporal location, attribute matching and other information, and associate the track with the target identity information.

[0076] Specifically, the detailed target location information (e.g., targets identified by photoelectric or SAR) is associated with existing tracks (track data), and the target's identity information (such as target number or category) is assigned to the matching track. The association criteria include spatiotemporal location and attribute matching, where spatiotemporal location refers to the proximity of the target and track in time and space; the degree of attribute matching refers to the degree of matching between the target and track's attribute information (such as type, size, etc.).

[0077] Step S9: Confirm the optical / SAR fusion attributes of the matched targets;

[0078] In step S9, multimodal slices, attributes, and identity category labels of targets associated with the same track are fused to confirm the target's military / civilian status, category, and other attributes, and to correct erroneous associated points / tracks. The optical / SAR fusion attribute confirmation is based on weighted voting and feature fusion. First, weighted voting is performed to determine the category of each feature. Then, combined with the multimodal feature information of the optical / SAR targets associated with the track, the specific implementation method for each step of the fusion algorithm is given. Finally, the track is matched with the fused target information.

[0079] Step S10: Associate the target trajectory with the existing situation map and catalog the target into the database;

[0080] Specifically, the target flight path is associated with the existing situation map, cataloged with a number, and stored in the database.

[0081] Step S11: Update and maintain the situation map based on the situation information recorded on the ground;

[0082] Specifically, based on the ground-based situation, the satellite incrementally updates and maintains the situation map using the ground-based situation as the situation anchor point, associating new tracks and points with targets in the original situation map, and adding new targets when there are unknown new tracks.

[0083] Step S12: Generate and distribute situation packets based on the situation map.

[0084] Specifically, at pre-set time intervals, or as needed through interpolation or short-term prediction, satellite-based distribution of sea surface situation packets for the required time points is generated, transmitted back to the ground, and broadcast. For point-to-point, layered, customized situation distribution, situation packets in the corresponding format are generated according to specific requirements and situation format templates and sent to the requesting party.

[0085] like Figure 2 and Figure 3As shown, this invention employs a containerized space-based heterogeneous multi-source information association and fusion device, comprising a track association and fusion function container, an attribute joint confirmation function container, and a situation generation function management container. The track association and fusion function container performs spatiotemporal alignment of the incoming decision-level information, real-time track-track association matching and fusion completion, and completes the batching and storage of tracks. The attribute joint confirmation function container performs point-track association matching and optical / SAR fusion attribute confirmation between the incoming optical / SAR information and the stored track information, and associates the targets with the confirmed fusion attributes with existing situation maps, and catalogs and stores them in the database. The situation generation function management container updates and maintains the situation map based on the ground-based situation report, and generates and distributes situation packages.

[0086] When performing space-based heterogeneous multi-source information association and fusion, the system first completes the access of track and optical / SAR decision-level information through the track association and fusion function container. Then, the tracks are spatiotemporally aligned, track-track association matching and fusion completion are implemented, and the tracks are batched and entered into the database. Next, the optical / SAR slices are calculated through the attribute joint confirmation function container, thereby completing the confirmation of optical / SAR fusion attributes and target cataloging. Finally, the situation map is updated and maintained through the situation generation function management container, and the situation package is generated and distributed.

[0087] In this invention, the containerized space-based heterogeneous multi-source information association and fusion device is deployed and implemented based on edge micro-clouds. Satellite clusters can be divided into micro-cloud nodes according to orbit or region. Lightweight K8s orchestration tools such as K3s / RKE2 are used for container orchestration to achieve "one cloud with multiple satellites and one satellite with multiple containers". This effectively improves the flexibility, scalability, reliability and maintainability of the space-based heterogeneous multi-source information association and fusion system, enabling it to better adapt to the dynamic changes in the space-based environment and the increasing mission requirements. This lays a solid technical foundation for future intelligent and autonomous upgrades.

[0088] Finally, it should be noted that the above description represents a preferred embodiment of the present invention. It should be pointed out that although preferred embodiments have been described, those skilled in the art, once they understand the basic inventive concept of the present invention, can make various improvements and modifications without departing from the principles described herein. These improvements and modifications should also be considered within the scope of protection of the present invention. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the embodiments of the present invention.

Claims

1. A containerized space-based heterogeneous multi-source information correlation and fusion method, characterized in that, Includes the following steps: Step S1: Analyze the trajectory data of the sea surface target, and store the acquired raw trajectory decision information into the database to complete the information access; Step S2: Perform spatiotemporal alignment of the acquired original track decision information; Step S3: Perform track-track association matching on the completed spatiotemporal alignment information; Step S4: Perform track fusion and completion based on the track matching results; Step S5: Batch the completed track fusion and data entry into the database. Step S6: Slice and analyze the optical / SAR of the sea surface target to complete the access of optical / SAR decision-level information; Step S7: Calculate the correlation between optical / SAR slices and identify target locations; Step S8: Based on the spatiotemporal location and attribute matching degree, perform point-track association matching between optical / SAR target location information and flight track; Step S9: Confirm the optical / SAR fusion attributes of the matched targets; Step S10: Associate the target trajectory with the existing situation map and catalog the target into the database; Step S11: Update and maintain the situation map based on the situation information recorded on the ground; Step S12: Generate and distribute situation packets based on the situation map.

2. The containerized space-based heterogeneous multi-source information correlation and fusion method according to claim 1, characterized in that, In the track-track association matching in step S3, Mahalanobis distance is used to measure the similarity of the state estimates of short tracks, the Hungarian algorithm is used to solve the many-to-many assignment problem to optimize the association, and a sequential pairwise association method is adopted to perform track association, ultimately achieving global track association.

3. The containerized space-based heterogeneous multi-source information correlation and fusion method according to claim 1, characterized in that, In step S4, the track fusion and completion includes the fusion operation of overlapping time periods and the completion operation of missing time periods. The fusion operation for the overlapping time period of the tracks includes weighted fusion of overlapping tracks associated with the same target; The missing time period completion operation includes interpolating or predicting the trajectory of the missing time period to complete it.

4. The containerized space-based heterogeneous multi-source information correlation and fusion method according to claim 1, characterized in that, Step S5 specifically includes: Each independent and complete track is assigned a unique batch number. Each independent and complete track includes the starting track, the bifurcation track, and the disappearing track. New tracks that cannot be associated with existing batch numbers are assigned new batch numbers.

5. The containerized space-based heterogeneous multi-source information correlation and fusion method according to claim 1, characterized in that, In step S7, the optical / SAR slice correlation calculation includes: calculating the correlation degree of the same modal slices or cross-modal multimodal slices that are initially screened based on spatiotemporal location under different viewing angles and different lighting conditions, to determine if they are the same target.

6. The containerized space-based heterogeneous multi-source information correlation and fusion method according to claim 1, characterized in that, Step S8 specifically includes: Point-track association matching involves associating the target location information identified by optical / SAR detailed investigation with existing tracks based on spatiotemporal location and attribute matching information, thus associating the tracks with target identity information. Spatiotemporal location refers to the proximity of the target and track in time and space; attribute matching degree refers to the degree of matching between the attribute information of the target and track.

7. The containerized space-based heterogeneous multi-source information correlation and fusion method according to claim 1, characterized in that, Step S11 specifically includes: Based on the ground-based situation, the situation map is updated and maintained incrementally using the ground-based situation as the situation anchor point, and the new flight paths and points are associated with the targets in the original situation map.

8. A containerized space-based heterogeneous multi-source information correlation and fusion device, used to implement the method as described in any one of claims 1 to 7, characterized in that, include: The trajectory association and fusion function container is used to perform spatiotemporal alignment of the accessed decision-level information, real-time trajectory-trajectory association matching and fusion completion, and complete the batch compilation and storage of trajectories. The attribute joint confirmation function container is used to perform point-track association matching and optical / SAR fusion attribute confirmation on the accessed optical / SAR information and the track information in the database, and to associate the target after fusion attribute confirmation with the existing situation map and catalog it into the database. The situation generation function management container is used to update and maintain situation maps based on ground-based situation data, and to generate and distribute situation packages.

9. The containerized space-based heterogeneous multi-source information correlation and fusion device according to claim 8, characterized in that, The containerized space-based heterogeneous multi-source information association and fusion device is deployed and implemented in a containerized manner based on edge micro-cloud.

Images