Distributed device for rapid BP imaging of space-borne SAR based on geographic coordinate system

Abstract

This invention provides a distributed device for fast backpropagation (BP) imaging of spaceborne SAR based on a geographic coordinate system, relating to the field of radar signal processing technology. It includes: multiple clustered slave nodes, used to acquire data corresponding to each sub-aperture from a distributed storage module; utilizing a fast BP imaging algorithm, sequentially processing all sub-aperture data in parallel within each pulse-compressed echo data block to obtain multiple sub-aperture images for each pulse-compressed echo data block; fusing these multiple sub-aperture images using an upsampling method to obtain multiple sub-scene images; and sending these multiple sub-scene images to a distributed aggregation module; the distributed aggregation module coherently sums these multiple sub-scene images to obtain at least one full-resolution single-scene image; and determining the imaging image based on the single-scene image. This results in shorter imaging processing time, higher imaging resolution, and lower imaging costs in high-orbit imaging.

Description

Technical Field

[0001] This invention relates to the field of radar signal processing technology, and in particular to a distributed spaceborne SAR fast BP imaging device based on a geographic coordinate system. Background Technology

[0002] With the increasing maturity of synthetic aperture radar (SAR) technology, it has high practical value and wide application scenarios in both defense and civilian fields. In recent years, SAR imaging has been developing towards ultra-high resolution and ultra-wide imaging swaths in high-orbit environments. Ultra-high resolution imaging and ultra-wide imaging swaths in high-orbit environments mean processing massive amounts of data and performing enormous calculations, which places high demands on both algorithms and computing hardware. The geographic coordinate system-based fast error backpropagation (BP) algorithm plays a crucial role in high-resolution and high-orbit applications. Its high imaging quality and lack of interpolation requirements make it suitable for engineering applications. Graphics Processing Units (GPUs), typically paired with Central Processing Units (CPUs), are computing devices that provide powerful computing capabilities. Initially, GPUs were mainly used for computer graphics and image processing tasks, such as games and video playback. With technological advancements, the application areas of GPUs have expanded to scientific computing, machine learning, and data processing. With the maturity and improvement of GPU technology, the SAR imaging field has also adopted GPUs as computing processing devices to process SAR data.

[0003] Currently, SAR imaging primarily utilizes the Fast Decomposition Backpropagation (BP) algorithm and employs CPUs and GPUs in its hardware architecture. CPU and GPUs can collaborate, with the CPU acting as the host, mainly responsible for preprocessing SAR data, such as pulse compression and grid construction. The processed SAR data is then distributed to the GPU for processing according to azimuth pulses. The GPU receives the pulse data, performs imaging processing, and sends the processed image back to the host. However, the BP algorithm requires interpolation, which can degrade image quality. Furthermore, the BP algorithm is unsuitable for high-orbit applications where the surface grid is curved, resulting in lower image resolution. Ultra-high resolution imaging and ultra-large-scale high-orbit imaging require massive computational power. For processing large amounts of data, CPU-GPU collaboration necessitates increasing the number of processors. However, the computational performance of CPU-GPU collaboration is limited by bandwidth; processing speed does not increase continuously with the increase in processor count, leading to longer imaging processing times. Increasing the number of processors also makes imaging too costly. Summary of the Invention

[0004] The purpose of this invention is to provide a distributed spaceborne SAR fast BP imaging device based on a geographic coordinate system, which solves the problems of low imaging resolution, long imaging processing time, and high imaging cost.

[0005] To address the aforementioned technical problems, the embodiments of the present invention provide the following technical solutions:

[0006] The first aspect of this invention provides a distributed spaceborne SAR fast BP imaging device based on a geographic coordinate system, the distributed spaceborne SAR fast BP imaging device based on a geographic coordinate system comprising:

[0007] The distributed resource scheduling module is used to divide the echo data into multiple echo data blocks in the azimuth direction according to the number of points in the azimuth direction, perform range pulse compression on the multiple echo data blocks to obtain multiple pulse-compressed echo data blocks, and take each pulse-compressed echo data block as an imaging sub-scene. According to the preset number of columns, each imaging sub-scene is divided into multiple sub-apertures, a cluster of working slave nodes is assigned to each sub-aperture, and the data corresponding to each sub-aperture is sent to the distributed storage module.

[0008] A distributed storage module is used to store the data corresponding to each sub-aperture;

[0009] Multiple clusters of slave nodes are used to retrieve data corresponding to each sub-aperture from the distributed storage module. Using the fast BP imaging algorithm, the data corresponding to all sub-apertures in each pulse-compressed echo data block are processed in parallel to obtain multiple sub-aperture images corresponding to each pulse-compressed echo data block. The multiple sub-aperture images are fused by the upsampling method to obtain multiple sub-scene images, and the multiple sub-scene images are sent to the distributed aggregation module.

[0010] The distributed aggregation module is used to coherently add multiple sub-scene images to obtain at least one full-resolution single-scene image, and determine the imaging image based on the single-scene image.

[0011] A second aspect of this invention provides a distributed method for fast BP imaging of spaceborne SAR based on a geographic coordinate system. The distributed method for fast BP imaging of spaceborne SAR based on a geographic coordinate system includes:

[0012] Based on the number of points in the azimuth direction, the echo data is divided into multiple echo data blocks in the azimuth direction;

[0013] Range pulse compression is performed on multiple echo data blocks to obtain multiple pulse-compressed echo data blocks;

[0014] Each pulse-compressed echo data block is treated as an imaging sub-scene, and each imaging sub-scene is divided into multiple sub-apertures according to a preset number of columns.

[0015] Data corresponding to each sub-aperture is acquired. Using the fast BP imaging algorithm, all data corresponding to sub-apertures in each pulse-compressed echo data block are processed in parallel to obtain multiple sub-aperture images corresponding to each pulse-compressed echo data block. The multiple sub-aperture images are then fused using an upsampling method to obtain multiple sub-scene images.

[0016] Coherently add multiple sub-scene images to obtain at least one full-resolution single-scene image;

[0017] The imaging image is determined based on the single scene image.

[0018] Compared to existing technologies, the distributed spaceborne SAR fast BP imaging device based on a geographic coordinate system provided by this invention includes a distributed resource scheduling module. This module divides the echo data into multiple echo data blocks in the azimuth direction according to the number of points. Range pulse compression is then performed on these multiple echo data blocks to obtain multiple pulse-compressed echo data blocks. Each pulse-compressed echo data block is treated as an imaging sub-scene. Each imaging sub-scene is further divided into multiple sub-apertures according to a preset number of columns. A cluster of working slave nodes is assigned to each sub-aperture, and the data corresponding to each sub-aperture is sent to the distributed storage module. The distributed storage module stores the data for each sub-aperture. The data corresponds to the aperture; multiple cluster slave nodes are used to retrieve the data corresponding to each sub-aperture from the distributed storage module. Using the fast backpropagation (BP) imaging algorithm, all sub-aperture data in each pulse-compressed echo data block are processed in parallel to obtain multiple sub-aperture images corresponding to each pulse-compressed echo data block. The multiple sub-aperture images are fused using an upsampling method to obtain multiple sub-scene images, which are then sent to the distributed aggregation module. The distributed aggregation module is used to coherently add the multiple sub-scene images to obtain at least one full-resolution single-scene image. Based on the single-scene image, the imaging image is determined. In this way, the data corresponding to all sub-apertures in each pulse-compressed echo data block are deployed to a distributed device, that is, to multiple cluster slave nodes that can process in parallel. The multiple cluster slave nodes use the geographic coordinate system fast BP imaging algorithm to process the data corresponding to all sub-apertures in each pulse-compressed echo data block in parallel. The multiple cluster slave nodes can process the data corresponding to multiple sub-apertures simultaneously, which shortens the imaging processing time in high-orbit imaging, even with a large amount of data. Using the geographic coordinate system fast BP imaging algorithm, a full-resolution single-scene image can be obtained, resulting in high imaging resolution. Since the multiple cluster slave nodes can process the data corresponding to multiple sub-apertures simultaneously, no additional processors are needed, thus reducing the imaging cost. Attached Figure Description

[0019] The above and other objects, features, and advantages of exemplary embodiments of the present invention will become readily apparent upon reading the following detailed description with reference to the accompanying drawings. In the drawings, several embodiments of the invention are illustrated by way of example and not limitation, with the same or corresponding reference numerals denoteing the same or corresponding parts, wherein:

[0020] Figure 1 A schematic diagram of a distributed spaceborne SAR fast BP imaging device based on geographic coordinates is shown.

[0021] Figure 2A flowchart illustrating a distributed method for fast BP imaging of spaceborne SAR based on geographic coordinates is shown. Detailed Implementation

[0022] Exemplary embodiments of the invention will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the invention are shown in the drawings, it should be understood that the invention can be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the invention and to fully convey the scope of the invention to those skilled in the art.

[0023] It should be noted that, unless otherwise stated, the technical or scientific terms used in this invention should have the ordinary meaning as understood by one of ordinary skill in the art.

[0024] The methods described in the embodiments of the present invention will be explained in detail below.

[0025] Figure 1 The schematic diagram illustrates the structure of a distributed spaceborne SAR fast BP imaging device based on a geographic coordinate system according to an embodiment of the present invention. See also... Figure 1 As shown, the distributed spaceborne SAR fast BP imaging device based on geographic coordinate system may include:

[0026] The distributed resource scheduling module 101 is used to divide the echo data into multiple echo data blocks in the azimuth direction according to the number of points in the azimuth direction, perform range pulse compression on the multiple echo data blocks to obtain multiple pulse-compressed echo data blocks, take each pulse-compressed echo data block as an imaging sub-scene, divide each imaging sub-scene into multiple sub-apertures according to a preset number of columns, assign a cluster of working slave nodes to each sub-aperture, and send the data corresponding to each sub-aperture to the distributed storage module;

[0027] Distributed storage module 102 is used to store the data corresponding to each sub-aperture;

[0028] The working node 103 of the multiple clusters is used to obtain the data corresponding to each sub-aperture from the distributed storage module. Using the fast BP imaging algorithm, the data corresponding to all sub-apertures in each echo data block after pulse compression are processed in parallel to obtain multiple sub-aperture images corresponding to each echo data block after pulse compression. The multiple sub-aperture images are fused by the upsampling method to obtain multiple sub-scene images, and the multiple sub-scene images are sent to the distributed aggregation module.

[0029] The distributed aggregation module 104 is used to coherently add multiple sub-scene images to obtain at least one full-resolution single-scene image, and determine the imaging image based on the single-scene image.

[0030] Specifically, the number of pulse-compressed echo data blocks is the same as the number of multiple echo data blocks, and the number of pulse-compressed echo data blocks is the same as the number of multiple imaging sub-scenes. If the preset echo data block size is 8192*8192, it is divided into sub-apertures, each sub-aperture being a power of 2, typically 512 columns per sub-aperture, meaning the preset number of columns can be 512. The multiple sub-aperture images corresponding to a single pulse-compressed echo data block are fused using an upsampling method to obtain a sub-scene image. The multiple sub-aperture images corresponding to multiple pulse-compressed echo data blocks are then fused separately to obtain multiple sub-scene images. These multiple sub-scene images are low-resolution images.

[0031] The distributed resource scheduling module, distributed storage module, worker slave nodes of multiple clusters, and distributed aggregation module are connected in sequence. The distributed resource scheduling module is used to manage the running status of worker slave nodes of multiple clusters and distribute tasks to worker slave nodes of each cluster, that is, to distribute the data corresponding to each sub-aperture, with the data corresponding to one sub-aperture being considered as a task.

[0032] The distributed storage module also stores the running status of the worker nodes in each cluster. Multiple worker nodes in the cluster process the data corresponding to each sub-aperture in parallel. The distributed aggregation module is also used to display images from multiple sub-scenes.

[0033] The distributed resource scheduling module can be one or two high-performance servers.

[0034] Before the distributed resource scheduling module, distributed storage module, multiple cluster worker nodes, and distributed aggregation module determine the imaging image, a distributed computing framework needs to be built. Specifically, this involves: using Zookeeper as the distributed coordination service tool to build a distributed environment and store the running status, load status, and task execution status of distributed nodes; using the Hadoop Distributed Filesystem (HDFS) as the distributed data storage environment; using Kafka as the message middleware in the distributed environment to handle communication between modules and push intermediate results; and using the Docker container engine to implement containerized deployment of services and container orchestration in the distributed environment.

[0035] A distributed resource scheduling module is established as the master node (Manager node). The Manager node is responsible for task scheduling and load balancing of the distributed device, managing the running status of multiple cluster worker nodes, distributing data (tasks) for each sub-aperture, managing task load status, and managing task progress. Multiple worker nodes (Worker nodes) also need to be established. There are P worker nodes in the multiple clusters, where P≥2. The specific number of worker nodes can be configured according to actual task requirements. The worker nodes are responsible for executing the tasks distributed by the master node, i.e., the data corresponding to each sub-aperture. After establishing the worker nodes, a distributed file system is established, divided into storage source nodes (Storage Source) and storage destination nodes (StorageDestination). There is one storage source node and multiple storage destination nodes.

[0036] Since the number of echo data blocks after pulse compression is large enough, the number of imaging sub-scenes is also large enough. Furthermore, the computation time for the data corresponding to each sub-aperture is much longer than the time for prefetching echo data blocks, which can effectively hide the transmission and computation time.

[0037] The master and slave nodes provided in this embodiment of the invention, together with the corresponding low-coupling algorithm, form a distributed computing framework that is applicable to echo data processing in various situations. The modules are fully decoupled, enabling dynamic scaling up and down of capacity, rapid data processing, rapid display of results, and dynamic task scheduling.

[0038] In this embodiment, the working slave node 103 of the multiple clusters is specifically used to obtain the data corresponding to all sub-apertures in the first pulse-compressed echo data block from the distributed storage module. Using the fast backpropagation (BP) imaging algorithm, the data corresponding to all sub-apertures in the first pulse-compressed echo data block is processed in parallel to obtain multiple first sub-aperture images corresponding to the first pulse-compressed echo data block. The multiple first sub-aperture images are then fused using an upsampling method to obtain the first sub-scene image. The m-th processing step involves obtaining the data corresponding to all sub-apertures in the m-th pulse-compressed echo data block from the distributed storage module, and using the fast backpropagation (BP) imaging algorithm to process the data corresponding to all sub-apertures in the m-th pulse-compressed echo data block... Parallel processing is performed to obtain multiple m-th sub-aperture images corresponding to the m-th pulse-compressed echo data block. These multiple m-th sub-aperture images are then fused using an upsampling method to obtain the m-th sub-scene image. It is then determined whether the m-th pulse-compressed echo data block is the last pulse-compressed echo data block. If so, the first and m-th sub-scene images are treated as multiple sub-scene images and sent to the distributed aggregation module. If not, the process returns to the m-th iteration until the m-th pulse-compressed echo data block becomes the last pulse-compressed echo data block. At this point, processing stops, and the first and m-th sub-scene images are treated as multiple sub-scene images and sent to the distributed aggregation module.

[0039] In this embodiment, the distributed aggregation module 104 determines the imaging image based on the single scene image, including: determining whether the single scene image is of a preset size; if so, using the single scene image as the imaging image; if not, stitching the single scene images together to output a full scene image, and using the full scene image as the imaging image.

[0040] Specifically, the preset size can be 8192*8192 or 4096*4096. There is no specific limitation on the preset size here. The preset size is determined according to the actual echo data.

[0041] In this embodiment, the spaceborne SAR fast BP imaging distributed device based on geographic coordinate system also includes a distributed file system, which includes a storage source node and multiple storage destination nodes.

[0042] In this embodiment, the distributed resource scheduling module is allocated storage source nodes, and multiple clusters are allocated multiple storage destination nodes from their working slave nodes. The storage source nodes are used to store echo data in a distance-oriented continuous manner, and the multiple storage destination nodes are used to store data corresponding to each sub-aperture.

[0043] In this embodiment, the number of multiple storage destination nodes is the same as the number of multiple cluster slave nodes, and the multiple storage destination nodes and the multiple cluster slave nodes are allocated in a one-to-one correspondence.

[0044] Specifically, when there are P working nodes in multiple clusters, where P ≥ 2, there are also P corresponding storage destination nodes, where P ≥ 2.

[0045] In this embodiment, the working slave node 103 of the multiple clusters is also used to send multiple sub-scene images to multiple storage destination nodes, and when there is an image query requirement, read the corresponding sub-scene image and send the read sub-scene image to the distributed aggregation module.

[0046] In this embodiment, the distributed aggregation module 104 is used to acquire the read sub-scene images and display the read sub-scene images.

[0047] Based on the above Figure 1As can be seen from the implementation method, the distributed resource scheduling module of this embodiment of the invention is used to divide the echo data into multiple echo data blocks in the azimuth direction according to the number of points in the azimuth direction, perform range pulse compression on the multiple echo data blocks to obtain multiple pulse-compressed echo data blocks, and take each pulse-compressed echo data block as an imaging sub-scene. According to a preset number of columns, each imaging sub-scene is divided into multiple sub-apertures, and a cluster of working slave nodes is assigned to each sub-aperture. The data corresponding to each sub-aperture is sent to the distributed storage module; the distributed storage module is used to store the data corresponding to each sub-aperture; multiple clusters The group's working nodes retrieve data corresponding to each sub-aperture from the distributed storage module. Using the fast backpropagation (BP) imaging algorithm, it processes all sub-aperture data in parallel within each pulse-compressed echo data block, obtaining multiple sub-aperture images for each pulse-compressed echo data block. These multiple sub-aperture images are then fused using an upsampling method to obtain multiple sub-scene images, which are then sent to the distributed aggregation module. The distributed aggregation module coherently adds the multiple sub-scene images to obtain at least one full-resolution single-scene image. Based on this single-scene image, the final image is determined. In this way, the data corresponding to all sub-apertures in each pulse-compressed echo data block are deployed to a distributed device, that is, to multiple cluster slave nodes that can process in parallel. The multiple cluster slave nodes use the geographic coordinate system fast BP imaging algorithm to process the data corresponding to all sub-apertures in each pulse-compressed echo data block in parallel. The multiple cluster slave nodes can process the data corresponding to multiple sub-apertures simultaneously, which shortens the imaging processing time in high-orbit imaging, even with a large amount of data. Using the geographic coordinate system fast BP imaging algorithm, a full-resolution single-scene image can be obtained, resulting in high imaging resolution. Since the multiple cluster slave nodes can process the data corresponding to multiple sub-apertures simultaneously, no additional processors are needed, thus reducing the imaging cost.

[0048] Based on the same inventive concept, as an implementation of the above-mentioned distributed device for fast BP imaging of spaceborne SAR based on geographic coordinate system, this embodiment of the invention also provides a distributed method for fast BP imaging of spaceborne SAR based on geographic coordinate system. Figure 2 The flowchart of the distributed method for fast BP imaging of spaceborne SAR based on geographic coordinate system in this embodiment of the invention is shown below. Figure 2 As shown, this distributed method for fast BP imaging of spaceborne SAR based on geographic coordinate systems may include:

[0049] S201. Based on the number of points in the azimuth direction, divide the echo data into multiple echo data blocks in the azimuth direction.

[0050] S202. Perform range pulse compression on multiple echo data blocks to obtain multiple pulse-compressed echo data blocks.

[0051] S203. Each pulse-compressed echo data block is treated as an imaging sub-scene, and each imaging sub-scene is divided into multiple sub-apertures according to a preset number of columns.

[0052] S204. Acquire the data corresponding to each sub-aperture. Using the fast BP imaging algorithm, process the data corresponding to all sub-apertures in each pulse-compressed echo data block in parallel to obtain multiple sub-aperture images corresponding to each pulse-compressed echo data block. Then, fuse the multiple sub-aperture images using an upsampling method to obtain multiple sub-scene images.

[0053] Specifically, data corresponding to each sub-aperture is acquired. Using a fast backpropagation (BP) imaging algorithm, the data corresponding to all sub-apertures in each pulse-compressed echo data block are processed in parallel to obtain multiple sub-aperture images for each pulse-compressed echo data block. These multiple sub-aperture images are then fused using an upsampling method to obtain multiple sub-scene images, including:

[0054] Step A1: Obtain the data corresponding to all sub-apertures in the first echo data block after pulse compression. Using the fast BP imaging algorithm, process the data corresponding to all sub-apertures in the first echo data block after pulse compression in parallel to obtain multiple first sub-aperture images corresponding to the first echo data block after pulse compression. Then, fuse the multiple first sub-aperture images using an upsampling method to obtain the first sub-scene image.

[0055] Step A2: m-th processing: Obtain the data corresponding to all sub-apertures in the m-th pulse-compressed echo data block. Using the fast BP imaging algorithm, process the data corresponding to all sub-apertures in the m-th pulse-compressed echo data block in parallel to obtain multiple m-th sub-aperture images corresponding to the m-th pulse-compressed echo data block. Then, fuse the multiple m-th sub-aperture images using an upsampling method to obtain the m-th sub-scene image.

[0056] Step A3: Determine whether the m-th pulse-compressed echo data block is the last pulse-compressed echo data block.

[0057] Specifically, determine whether the m-th pulse-compressed echo data block is the last pulse-compressed echo data block. If yes, proceed to step A4; otherwise, proceed to step A5.

[0058] Step A4: Treat the first sub-scene image and the m-th sub-scene image as multiple sub-scene images.

[0059] Specifically, when the m-th echo data block after pulse compression is the last echo data block after pulse compression, the first sub-scene image and the m-th sub-scene image are treated as multiple sub-scene images.

[0060] Step A5: Return to the m-th processing step until the m-th pulse-compressed echo data block becomes the last pulse-compressed echo data block, then stop processing and treat the first sub-scene image and the m-th sub-scene image as multiple sub-scene images.

[0061] Specifically, if the m-th pulse-compressed echo data block is not the last pulse-compressed echo data block, return to the m-th processing step, i.e., return to step A2, until the m-th pulse-compressed echo data block is the last pulse-compressed echo data block, then stop processing and treat the first sub-scene image and the m-th sub-scene image as multiple sub-scene images.

[0062] S205. Coherently add multiple sub-scene images to obtain at least one full-resolution single-scene image.

[0063] S206. Determine the image based on the single-scene image.

[0064] Specifically, based on a single scene image, the resulting image is determined, including:

[0065] Step B1: Determine if the single-scene image is the preset size.

[0066] Specifically, determine whether the single-scene image is of the preset size. If yes, proceed to step B2; otherwise, proceed to step B3.

[0067] Step B2: Use the single-scene image as the imaging image.

[0068] Specifically, when the single-view image is of a preset size, the single-view image is used as the imaging image.

[0069] Step B3: Stitch together the single-scene images to output the full-scene image, and use the full-scene image as the imaging image.

[0070] Specifically, when a single scene image is not the preset size, the single scene images are stitched together to output a full-scene image, which is then used as the imaging image.

[0071] It should be noted that the above description of the distributed method embodiment for spaceborne SAR fast BP imaging based on geographic coordinates is similar to the description of the distributed device embodiment for spaceborne SAR fast BP imaging based on geographic coordinates, and has similar beneficial effects. For technical details not disclosed in the embodiments of the distributed method for spaceborne SAR fast BP imaging based on geographic coordinates of the present invention, please refer to the description of the distributed device embodiment for spaceborne SAR fast BP imaging based on geographic coordinates of the present invention for understanding.

[0072] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A kind of satellite-borne SAR fast BP imaging distributed device based on geographic coordinate system, characterized by, The distributed spaceborne SAR fast BP imaging device based on geographic coordinate system includes: The distributed resource scheduling module is used to divide the echo data into multiple echo data blocks in the azimuth direction according to the number of points in the azimuth direction, perform range pulse compression on the multiple echo data blocks to obtain multiple pulse-compressed echo data blocks, take each pulse-compressed echo data block as an imaging sub-scene, divide each imaging sub-scene into multiple sub-apertures according to a preset number of columns, assign a cluster of working slave nodes to each sub-aperture, and send the data corresponding to each sub-aperture to the distributed storage module; The distributed storage module is used to store the data corresponding to each sub-aperture; Multiple clusters of slave nodes are used to obtain data corresponding to each sub-aperture from the distributed storage module. Using the fast BP imaging algorithm, the data corresponding to all sub-apertures in each pulse-compressed echo data block are processed in parallel to obtain multiple sub-aperture images corresponding to each pulse-compressed echo data block. The multiple sub-aperture images are fused by an upsampling method to obtain multiple sub-scene images, and the multiple sub-scene images are sent to the distributed aggregation module. The distributed aggregation module is used to coherently add the multiple sub-scene images to obtain at least one full-resolution single-scene image, and determine the imaging image based on the single-scene image. The distributed device for spaceborne SAR fast BP imaging based on geographic coordinate system also includes a distributed file system, which includes a storage source node and multiple storage destination nodes. The distributed resource scheduling module is allocated the storage source node, and the multiple clusters are allocated the multiple storage destination nodes from the working slave nodes. The storage source node is used to store the echo data in a distance-oriented continuous manner, and the multiple storage destination nodes are used to store the data corresponding to each sub-aperture. The number of the plurality of storage destination nodes is the same as the number of the plurality of cluster slave nodes, and the plurality of storage destination nodes and the plurality of cluster slave nodes are allocated in a one-to-one correspondence; The distributed spaceborne SAR fast BP imaging device based on geographic coordinate system adopts a distributed device computing framework, which includes distributed application coordination service software, the distributed file system, Kafka, and application container engine. The distributed resource scheduling module is used as the master node. The master node is responsible for task scheduling and load balancing of the distributed device, managing the running status of the slave nodes of the multiple clusters, distributing tasks corresponding to each sub-aperture, managing task load status and task progress. The working slave nodes of the multiple clusters are also used to send the multiple sub-scene images to the multiple storage destination nodes, and when there is an image query requirement, read the corresponding sub-scene image and send the read sub-scene image to the distributed aggregation module.

2. The distributed device for fast space-borne SAR BP imaging based on geographic coordinate system according to claim 1, wherein, The multiple clusters of slave nodes are specifically used to obtain data corresponding to all sub-apertures in the first pulse-compressed echo data block from the distributed storage module, and to process the data corresponding to all sub-apertures in the first pulse-compressed echo data block in parallel using a fast backpropagation (BP) imaging algorithm, thereby obtaining multiple first sub-aperture images corresponding to the first pulse-compressed echo data block. These multiple first sub-aperture images are then fused using an upsampling method to obtain the first sub-scene image. Second processing: Obtain the first... Data corresponding to all sub-apertures in the echo data block after pulse compression are analyzed using the fast BP imaging algorithm. The data corresponding to all sub-apertures in the pulse-compressed echo data block are processed in parallel to obtain the data of the first... The multiple echo data blocks corresponding to the pulse compression Sub-aperture images, which are then upsampled using a method to capture the multiple first aperture images. Sub-aperture images are fused to obtain the first The first sub-scene image; determine the first If the first pulse-compressed echo data block is the last pulse-compressed echo data block, then the first sub-scene image and the second... If one sub-scene image is used as the plurality of sub-scene images, the plurality of sub-scene images are sent to the distributed aggregation module; otherwise, the process returns to the previous step. The process continues until the [number]th [process]. When the echo data block after pulse compression is the last echo data block after pulse compression, processing stops, and the first sub-scene image and the second... Each sub-scene image is treated as a plurality of sub-scene images, and the plurality of sub-scene images are sent to the distributed aggregation module. 3.The fast BP imaging of space-borne SAR based on geographic coordinate system distributed device according to claim 1, wherein, The distributed aggregation module determines the imaging image based on the single-scene image, including: determining whether the single-scene image is of a preset size; if so, using the single-scene image as the imaging image; otherwise, stitching the single-scene images together to output a full-scene image, and using the full-scene image as the imaging image.

4. The distributed device for fast space-borne SAR BP imaging based on geographic coordinate system according to claim 1, wherein, The distributed aggregation module is used to acquire the read sub-scene images and display them.

5. A distributed method for fast BP imaging of space-borne SAR based on geographic coordinate system, characterized in that, The distributed method for spaceborne SAR fast BP imaging based on geographic coordinate system, applicable to any one of claims 1-4, comprises: Based on the number of points in the azimuth direction, the echo data is divided into multiple echo data blocks in the azimuth direction; Range pulse compression is performed on the plurality of echo data blocks to obtain a plurality of pulse-compressed echo data blocks; Each pulse-compressed echo data block is treated as an imaging sub-scene, and each imaging sub-scene is divided into multiple sub-apertures according to a preset number of columns. Data corresponding to each sub-aperture is acquired. Using the fast BP imaging algorithm, all data corresponding to sub-apertures in each echo data block after pulse compression are processed in parallel to obtain multiple sub-aperture images corresponding to each echo data block after pulse compression. The multiple sub-aperture images are then fused using an upsampling method to obtain multiple sub-scene images. The multiple sub-scene images are coherently added together to obtain at least one full-resolution single-scene image; Based on the single-scene image, determine the imaging image; Before acquiring the data corresponding to each sub-aperture, the method further includes: storing the echo data in a distance-continuous manner, and storing the data corresponding to each sub-aperture.

6. The distributed method of space-borne SAR fast BP imaging based on geographic coordinate system according to claim 5, characterized in that, The step of determining the image based on the single-scene image includes: Determine whether the single-scene image is of a preset size; If so, the single-scene image is used as the imaging image; If not, the single-scene images are stitched together to output a full-scene image, which is then used as the imaging image.

Images