Cloud layer interference resistant target detection method and system based on multi-satellite on-orbit scheduling

Abstract

This application provides a target detection method and system based on multi-satellite on-orbit scheduling to combat cloud interference. The method includes: deploying a satellite constellation architecture, which includes a central satellite and multiple edge satellites; capturing images of the area to be detected to obtain initial remote sensing images, and identifying and classifying cloud thickness in the initial remote sensing images to determine thin cloud regions and thick cloud regions; identifying at least one area to be captured from the thick cloud-covered areas corresponding to the thick cloud region images; generating a capture schedule table for the areas to be captured; capturing images of the areas to be captured using the target edge satellites according to the capture schedule table to obtain target images of the areas to be captured; performing target detection on the thin cloud region images and the target images to determine the targets in the areas to be detected. By optimizing the scheduling and cooperation between satellites, the method can effectively capture targets in the areas to be detected.

Description

Technical Field

[0001] This application relates to the field of satellite computing technology, and in particular to a method and system for target detection against cloud interference based on multi-satellite on-orbit scheduling. Background Technology

[0002] The primary application of most optical remote sensing images is target detection, aiming to identify and extract specific ground features from surface images. However, in practical applications, approximately 67% of remote sensing images are affected by cloud cover, resulting in a decline in image quality. This cloud contamination not only obscures ground targets but also alters the image's texture, brightness, and contrast, significantly increasing the difficulty of target detection. Furthermore, the complex morphology of clouds makes it even more challenging for detection algorithms to process these cloud-affected images. Therefore, effectively addressing cloud interference remains a crucial research issue in optical remote sensing image processing for satellite-to-ground target detection.

[0003] In related technologies, the following methods are typically used to deal with cloud interference:

[0004] 1. Satellite "bend tube" transmission transmits all image data back to the ground station, where cloud removal processing is performed on the images.

[0005] 2. Multi-time shooting: Take multiple shots at different times, select cloudless images, and stitch them together.

[0006] 3. The image was taken when a single satellite detected a cloudless period.

[0007] Based on the above methods, the following technical problems exist:

[0008] 1. "Bent pipe" transmission mode

[0009] In this mode, the satellite transmits all captured image data directly back to the ground station, where the data is then processed to remove cloud cover.

[0010] shortcoming:

[0011] 1) Wasting valuable inter-satellite link bandwidth resources:

[0012] Because satellites need to transmit all captured image data, including data affected by cloud cover, this "bent-pipe" transmission mode consumes a significant amount of bandwidth in the limited space-to-ground communication link. The transmission of large amounts of redundant data is not only inefficient but also wastes link resources, impacting the data transmission needs of other high-priority tasks.

[0013] 2) Increased processing burden on ground stations:

[0014] After receiving a large number of remote sensing images, the ground station needs to perform complex cloud removal processing on each image, especially in areas with thick clouds, where the computational cost is high and the processing time is long. This centralized processing mode not only delays the actual utilization efficiency of the data, but also significantly increases the processing workload of the ground station, placing higher demands on the hardware resources and computing power of the ground station.

[0015] 2. Multi-time shooting mode

[0016] Multi-temporal shooting mode attempts to select cloudless image areas from multiple shots of the same area at different times, and then synthesize high-quality remote sensing images.

[0017] shortcoming:

[0018] 1) Poor timeliness:

[0019] Multi-phase imaging requires waiting for satellites to pass over the same area before another image can be captured. In Low Earth Orbit (LEO) satellites, a single satellite typically needs an interval of about 90 minutes before it can reach the same target area again. If the target area is not continuously covered across multiple orbits, the timeliness of this mode is further reduced, failing to meet the demand for high-frequency and real-time remote sensing data, especially in emergency situations.

[0020] 2) Poor accuracy:

[0021] The dynamic changes of clouds are highly unpredictable. Even with multiple shots capturing images at different times, it is difficult to guarantee that subsequent shots will capture cloudless images of the target area. This is especially true in monsoon regions or under frequent cloudy weather conditions, where the reliability of multi-time-phase shooting is even lower, making it difficult for synthesized images to meet application requirements in terms of sharpness and coverage.

[0022] 3. Photographed when a single satellite detects no clouds.

[0023] This mode relies on satellites carrying cloud detectors to detect the distribution of clouds in the target area in real time. Once a cloudless area is detected, the satellite then takes precise pictures.

[0024] shortcoming:

[0025] 1) The detector is heavy and has high operating costs:

[0026] Cloud detectors are typically highly specialized equipment, increasing the weight and complexity of satellites. This not only places higher demands on satellite design and launch costs but also increases operational energy consumption and maintenance costs, limiting the widespread adoption of miniaturized, low-cost remote sensing satellites.

[0027] 2) Limitations of the shooting perspective and angle of a single satellite:

[0028] Even with cloud detectors, a single satellite is still limited by its orbit and camera angle when taking pictures, making it difficult to achieve flexible and wide coverage. At the same time, when the area obscured by thick clouds is large, relying solely on a single satellite to adjust the shooting angle to avoid obstruction has limited effectiveness, and may result in the target area not being fully covered. Summary of the Invention

[0029] In view of this, embodiments of this application provide a target detection method and system based on multi-satellite on-orbit scheduling to overcome or at least partially solve the above problems.

[0030] The first aspect of this application provides a cloud-resistant target detection method based on multi-satellite on-orbit scheduling, the method comprising:

[0031] Deploy a satellite constellation architecture, which includes a central satellite and multiple edge satellites;

[0032] The central satellite takes pictures of the area to be detected to obtain initial remote sensing images, and identifies and classifies the cloud thickness in the initial remote sensing images to determine the thin cloud area images and thick cloud area images in the initial remote sensing images.

[0033] Using the central satellite, at least one area to be photographed is determined from the thick cloud-covered area corresponding to the thick cloud area image;

[0034] Using the central satellite, a shooting schedule is generated for the area to be photographed. The shooting schedule is used to indicate the target edge satellite, the target observation window, and the target shooting time for photographing the area to be photographed. The target edge satellite is one of the multiple edge satellites.

[0035] According to the shooting schedule, the target area is photographed by the target edge satellite to obtain the target image of the target area;

[0036] Target detection is performed on the thin cloud region image and the target image to determine the target in the region to be detected.

[0037] Optionally, determining at least one area to be photographed from the thick cloud-covered area corresponding to the thick cloud area image includes:

[0038] For the image of the thick cloud region, the central satellite determines the thick cloud-covered area based on the latitude and longitude information of the thick cloud region image in the initial remote sensing image. The thick cloud-covered area is a rectangular area of ​​latitude and longitude of the thick cloud-obscured area in the image of the thick cloud region.

[0039] The minimum shooting angle for the thick cloud-covered area is determined based on the width of the thick cloud-covered area and the height of the cloud layer.

[0040] The area to be photographed is determined based on the minimum shooting angle.

[0041] Optionally, determining the area to be photographed based on the minimum shooting angle includes:

[0042] When the minimum shooting angle is not greater than the maximum side angle of the satellite, the thick cloud-covered area is determined as the area to be shot, and the minimum shooting angle of the area to be shot is the minimum shooting angle of the thick cloud-covered area.

[0043] When the minimum shooting angle is greater than the maximum side angle of the satellite, the thick cloud-covered area is divided into multiple shooting areas with the aim that the minimum shooting angle of the area to be shot is not greater than the maximum side angle of the satellite. The minimum shooting angle of the area to be shot is determined based on the width of the area to be shot and the height of the cloud layer.

[0044] The method further includes:

[0045] When the thick cloud-covered area is the area to be photographed, an edge satellite is identified as the target edge satellite, and the area to be photographed is matched with an edge satellite;

[0046] When the thick cloud-covered area is divided into multiple areas to be photographed, multiple edge satellites are identified as target edge satellites, and multiple areas to be photographed are matched with multiple edge satellites respectively.

[0047] Optionally, before generating the shooting schedule for the area to be shot, the method further includes:

[0048] For each of the multiple edge satellites, obtain the shooting angle change information and time interval of the edge satellite; the shooting angle change information includes: the shooting angle of the edge satellite at the last shooting and the minimum shooting angle of the area to be shot; the time interval is: the time required for the edge satellite to turn from the current shooting angle to the maximum side angle of the satellite, and the current shooting angle is the angle of the camera of the edge satellite when the area to be shot is determined;

[0049] The shooting angle torsion speed of the edge satellite is determined based on the shooting angle change information and the time interval.

[0050] Multiple edge satellites whose shooting angle torsion speed is not greater than the maximum speed of the satellite's camera angle torsion are selected as initial edge satellites;

[0051] The target edge satellite and the target observation window are determined from the initially selected edge satellites;

[0052] The target acquisition time is determined based on the current position of the target edge satellite;

[0053] The target observation window is determined based on the available observation window of the target edge satellite; the available observation window of the target edge satellite represents the range of the target edge satellite's shooting angle from the minimum shooting angle of the area to be photographed to the maximum side sway angle of the satellite.

[0054] Optionally, the process of selecting multiple edge satellites whose shooting angle twist speed is not greater than the maximum speed of the satellite's camera angle twist speed as initial edge satellites based on the shooting angle change information is as follows:

[0055] ;

[0056] in, This indicates the shooting angle of the last time the edge satellite was photographed. The minimum shooting angle of the area to be photographed; Indicates the time interval; This indicates the maximum speed at which the camera angle of the satellite twists.

[0057] Optionally, determining the target edge satellite from the initially selected edge satellites includes:

[0058] When there are multiple initial edge satellites, the available observation window of the initial edge satellite is obtained; the available observation window of the initial edge satellite represents the range from the first shooting angle of the initial edge satellite to the maximum side swing angle of the satellite when shooting the area to be shot.

[0059] Select the final edge satellites, which are the initial edge satellites with the largest available observation window, and use them as the target edge satellites.

[0060] Optionally, after the final selection of edge satellites, the process further includes:

[0061] When there are multiple final edge satellites, for each final edge satellite, obtain the shooting angle of the last shooting of the final edge satellite and the minimum shooting angle of the area to be shot;

[0062] The final edge satellite with the smallest camera twist angle between the shooting angle of the last shooting of the final edge satellite and the minimum shooting angle of the area to be photographed is selected as the target edge satellite.

[0063] Optionally, the plurality of edge satellites includes a first edge satellite group and a second edge satellite group; each edge satellite in the first edge satellite group operates along the same orbit as the central satellite, and each edge satellite in the second edge satellite group operates along an orbit with an inclination angle to the orbit of the central satellite;

[0064] The satellite constellation architecture is a satellite array centered on the central satellite, formed by a first edge satellite group and a second edge satellite group. The satellite constellation architecture is used to observe areas covered by thick clouds from multiple angles.

[0065] Optionally, the step of performing target detection on the thin cloud region image and the target image to determine the target in the region to be detected includes:

[0066] The image of the thin cloud area is processed by the central satellite to remove clouds.

[0067] Target detection is performed on the thin cloud region image after cloud removal processing to determine the target in the thin cloud region image. The image region where the target is located in the thin cloud region image is segmented to obtain the first target region image.

[0068] Target detection is performed on the target image to identify the target in the target image, and the image region where the target is located in the target image is segmented to obtain a second target region image;

[0069] The method further includes: sending the first target area image and the second target area image to the ground station through multiple transmission queues respectively.

[0070] The second aspect of this application provides a cloud-resistant target detection system based on multi-satellite on-orbit scheduling. The system includes a satellite constellation architecture and a ground station. The satellite constellation architecture includes a central satellite and multiple edge satellites.

[0071] The central satellite is used to take pictures of the area to be detected, obtain initial remote sensing images, and identify and classify the cloud thickness in the initial remote sensing images to determine the thin cloud area images and thick cloud area images in the initial remote sensing images.

[0072] The central satellite is used to determine at least one area to be photographed from the thick cloud-covered area corresponding to the thick cloud area image;

[0073] The central satellite is used to generate a shooting schedule for the area to be photographed. The shooting schedule is used to indicate the target edge satellite, the target observation window, and the target shooting time for photographing the area to be photographed. The target edge satellite is one of the multiple edge satellites.

[0074] The target edge satellite is used to take pictures of the area to be photographed according to the shooting schedule to obtain a target image of the area to be photographed.

[0075] The central satellite is used to perform target detection on the thin cloud region image and determine the target in the region to be detected;

[0076] The target edge satellite is used to perform target detection on the target image and determine the target in the area to be detected.

[0077] The beneficial effects of this application are:

[0078] This application provides a method and system for target detection against cloud interference based on multi-satellite on-orbit scheduling. The method includes: deploying a satellite constellation architecture, which includes a central satellite and multiple edge satellites; taking pictures of the area to be detected using the central satellite to obtain an initial remote sensing image, and identifying and classifying the cloud thickness in the initial remote sensing image to determine thin cloud region images and thick cloud region images in the initial remote sensing image; determining at least one area to be photographed from the thick cloud coverage area corresponding to the thick cloud region image; generating a shooting schedule table for the area to be photographed, the shooting schedule table indicating the target edge satellite, the target observation window, and the target shooting time for photographing the area to be photographed, wherein the target edge satellite is one of the multiple edge satellites; taking pictures of the area to be photographed using the target edge satellite according to the shooting schedule table to obtain a target image of the area to be photographed; and performing target detection on the thin cloud region image and the target image to determine the target in the area to be detected.

[0079] The technical solution provided in this application enables the effective capture of targets in areas under the cloud cover when thick cloud interference makes it difficult to effectively cover the ground surface. By optimizing the scheduling and cooperation among multiple satellites, the accuracy of the shooting effect and the image quality can be ensured. Attached Figure Description

[0080] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments of this application and their descriptions are used to explain this application and do not constitute an undue limitation of this application.

[0081] To more clearly illustrate the technical solution of this application, the drawings used in the description of this application will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0082] Figure 1This is a flowchart illustrating an embodiment of a cloud-based target detection method based on multi-satellite on-orbit scheduling;

[0083] Figure 2 This is a schematic diagram illustrating the deployment of a satellite constellation architecture according to an embodiment of this application;

[0084] Figure 3 This is a schematic diagram of the orbital design of the satellite constellation architecture shown in this application;

[0085] Figure 4 This is an overall workflow diagram of the satellite constellation architecture shown in this application;

[0086] Figure 5 This is a schematic diagram illustrating the calculation of the minimum shooting angle shown in this application;

[0087] Figure 6 This is a schematic diagram illustrating the maximum lateral sway angle shown in this application;

[0088] Figure 7 This is a schematic diagram illustrating the segmentation of a thick cloud-covered area as shown in this application;

[0089] Figure 8 This is a schematic diagram showing the available observation windows as illustrated in this application;

[0090] Figure 9 This is a schematic diagram illustrating the first stage of cloud removal and target recognition process shown in this application;

[0091] Figure 10 This is a schematic diagram illustrating the second stage of the target identification process shown in this application;

[0092] Figure 11 This is a schematic diagram illustrating the priority relationship of multi-level constraints as shown in this application. Detailed Implementation

[0093] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other.

[0094] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0095] refer to Figure 1 , Figure 1This is a flowchart illustrating a target detection method against cloud interference based on multi-satellite on-orbit scheduling, as shown in one embodiment of this application. Specifically, one embodiment of this application provides a target detection method against cloud interference based on multi-satellite on-orbit scheduling, which includes steps S11 to S16:

[0096] Step S11: Deploy a satellite constellation architecture, which includes a central satellite and multiple edge satellites.

[0097] This embodiment proposes an innovative multi-satellite-based imaging architecture (i.e., a satellite constellation architecture) to address the problem of cloud interference in optical remote sensing images during target detection tasks. This satellite constellation architecture consists of two parts: a central satellite and edge satellites. The central satellite consists of one satellite, while the edge satellites consist of six satellites distributed around the central satellite to enhance observation coverage and flexibility. A schematic diagram of the satellite constellation architecture is shown below. Figure 2 As shown.

[0098] Optionally, Figure 3 This is a schematic diagram of the orbital design of the satellite constellation architecture shown in this application, wherein multiple edge satellites include a first edge satellite group and a second edge satellite group. Each edge satellite in the first edge satellite group runs along the same orbit as the central satellite, and each edge satellite in the second edge satellite group runs along an orbit with an inclination angle to the orbit of the central satellite.

[0099] The satellite constellation architecture consists of a central satellite array formed by the first and second edge satellite groups. This constellation architecture is used to observe areas covered by thick clouds from multiple angles.

[0100] refer to Figure 3 The distribution of multiple edge-stars ensures that the satellite array can observe areas covered by thick clouds from six angles centered on the center-star, thus effectively avoiding the inability to meet the angle requirements for shooting cloudless areas and ensuring high-quality images can be obtained even when facing thick cloud conditions.

[0101] Effective geometric configuration enhances the relative positional diversity and observation angle of the satellite array under different observation conditions, thereby strengthening the system's resistance to cloud cover and improving the coverage of remote sensing images, thus effectively and minimizing the interference of cloud obstruction on image quality.

[0102] Step S12: The central satellite takes pictures of the area to be detected to obtain an initial remote sensing image, and identifies and classifies the cloud thickness in the initial remote sensing image to determine the thin cloud area image and the thick cloud area image in the initial remote sensing image.

[0103] Figure 4 This application illustrates the overall workflow of the satellite constellation architecture, as shown below. Figure 4 As shown, in this embodiment, the central satellite, as the core of the entire satellite constellation architecture, plays a command and coordination role similar to a "brain." It can be responsible for capturing initial remote sensing images of the area to be detected, and identifying and classifying the initial remote sensing images according to cloud regions of different thicknesses, obtaining images of thin cloud regions and images of thick cloud regions.

[0104] The area to be detected refers to the object of the target detection task performed by the central satellite, such as a certain surface area. Thin cloud area image refers to the area image in the initial remote sensing image where the cloud thickness is lower than the preset cloud thickness threshold, and thick cloud area image refers to the area image in the initial remote sensing image where the cloud thickness is not lower than the preset cloud thickness threshold. The preset cloud thickness threshold can be set according to whether the cloud thickness in the area image can be directly declouded.

[0105] Step S13: Using the central satellite, at least one area to be photographed is determined from the thick cloud coverage area corresponding to the thick cloud area image.

[0106] In this embodiment, the thick cloud-covered area refers to the specific coverage area corresponding to the thick cloud region image in the target detection area in the initial remote sensing image. However, due to the excessive cloud thickness in the thick cloud region image, it cannot be directly de-clouded before target detection. Therefore, to solve this problem, it is necessary to re-encode the thick cloud-covered area corresponding to the thick cloud region image using edge satellites. During the encoder, the edge satellites need to take oblique shots along the cloud edge to avoid directly aiming at the top of the cloud layer, thus obtaining an image of the region unaffected by cloud interference. This method of scheduling edge satellites for oblique shooting requires one or more edge satellites. Therefore, according to the needs of the shooting task, at least one target area needs to be determined from the thick cloud-covered area. Each edge satellite's shooting task corresponds to one target area. When the thick cloud-covered area is too large to be fully captured by a single edge satellite, the thick cloud-covered area needs to be divided into multiple target areas, and the shooting process is performed by multiple edge satellites.

[0107] Step S14: Using the central satellite, a shooting schedule is generated for the area to be photographed. The shooting schedule is used to indicate the target edge satellite, the target observation window, and the target shooting time for photographing the area to be photographed. The target edge satellite is one of the multiple edge satellites.

[0108] In this embodiment, for each area to be photographed, the central satellite will perform precise scheduling calculations based on the location of thick cloud cover in the area to be photographed, the status of the satellite camera, and the real-time position of the satellite. The shooting schedule will list the observation windows and shooting times of the edge satellites in detail to ensure that each edge satellite can perform effective target detection in the area covered by thick clouds.

[0109] Specifically, by accessing the current position and status information of each edge satellite, and combining the orbital data of each edge satellite, camera configuration, thick cloud cover areas, and constraints of shooting conditions, the central satellite generates a shooting schedule through complex mathematical modeling and algorithm analysis. This shooting schedule determines the edge satellites (target edge satellites) to be used for shooting the area, the shooting time for these target edge satellites (target shooting time), and the shooting angle range for these target edge satellites (target observation window). Based on this information, a shooting schedule for the area to be photographed is generated.

[0110] Step S15: According to the shooting schedule, the target area is photographed by the target edge satellite to obtain the target image of the target area.

[0111] Step S16: Perform target detection on the thin cloud region image and the target image to determine the target in the region to be detected.

[0112] Through the above embodiments, this application proposes a multi-satellite collaborative scheduling-based imaging strategy, which can optimize the scheduling and cooperation between satellites, enabling each satellite to effectively capture images of the area to be photographed within different observation windows. This maximizes the effectiveness of satellite image data, ensuring that all images transmitted back to the ground station are high-quality and usable data, and avoiding the large amount of bandwidth waste commonly found in traditional "bend-pipe" data transmission.

[0113] Furthermore, unlike traditional multi-temporal imaging methods, this application uses precise mathematical modeling and optimization algorithms to dynamically calculate the optimal imaging angle for edge satellites at specific times, generating an imaging schedule. This fully considers factors such as the location of areas covered by thick clouds, the satellite's position, and cloud conditions, ensuring that each satellite can acquire images at the optimal imaging angle. This compensates for the low real-time performance and timeliness, as well as the instability of image quality, caused by multi-temporal imaging.

[0114] In addition, compared to traditional single-satellite independent cloud-avoidance imaging missions, the multi-satellite collaborative approach in this application eliminates the need for complex additional hardware, significantly reducing system complexity and cost. Secondly, the real-time performance of multi-satellite imaging is greatly improved. Multiple satellites can simultaneously observe within different observation windows, and this collaborative approach significantly shortens the overall imaging time. Compared to single-satellite imaging, it can more efficiently cover the area to be photographed, especially in dynamically changing environments, enabling higher-frequency image acquisition.

[0115] Optionally, in conjunction with the above embodiments, one embodiment of this application also provides another target detection method based on multi-satellite on-orbit scheduling to resist cloud interference. In this method, the step S13 of "determining at least one area to be photographed from the thick cloud coverage area corresponding to the thick cloud area image" specifically includes steps S13-1 to S13-3:

[0116] Step S13-1: For the thick cloud region image, the central satellite determines the thick cloud coverage area based on the latitude and longitude information of the thick cloud region image in the initial remote sensing image. The thick cloud coverage area is a rectangular area of ​​latitude and longitude of the thick cloud occlusion area in the thick cloud region image.

[0117] In this embodiment, the determination of the thick cloud coverage area corresponding to the thick cloud area image in the area to be detected can be achieved by calculating the precise latitude and longitude rectangle of the part of the area to be detected that is obscured by thick clouds based on the latitude and longitude information of the four corners of the thick cloud area image collected in the initial remote sensing image, thus obtaining the thick cloud coverage area.

[0118] Step S13-2: Determine the minimum shooting angle for the thick cloud-covered area based on the width of the thick cloud-covered area and the height of the cloud layer.

[0119] like Figure 5 The diagram illustrates the principle of edge satellites capturing images of areas covered by thick clouds. In this embodiment, when the central satellite's imagery is obstructed by thick clouds, if an edge satellite located to the right of the central satellite is invoked for imaging, the edge satellite needs to establish a geometric relationship based on the left edge of the cloud-covered area and the right edge of the cloud layer to determine the most suitable shooting angle. Based on this most suitable shooting angle, a minimum shooting angle condition is set. The minimum shooting angle represents the smallest shooting angle that can completely capture the cloud-covered area. The most suitable shooting angle should not be less than the minimum shooting angle.

[0120] Figure 5 This is a schematic diagram illustrating the calculation of the minimum shooting angle shown in this application. For details, please refer to... Figure 5, this minimum shooting angle can be obtained through geometric calculation of the width (L) of the thick cloud coverage area and the cloud height (heightc) of the thick cloud coverage area, as shown by the following formula:

[0121] ;

[0122] Among them, the calculation formula of the minimum shooting angle is based on the width and cloud height of the marked thick cloud coverage area. Through the above formula, the required minimum shooting angle can be deduced. represents the minimum shooting angle.

[0123] During the shooting process of the remote sensing satellite, there is a significant difference between the cloud height (heightc) and the orbital height (heights) of the low Earth orbit (LEO) satellite. The typical height of clouds usually ranges between 2 and 12 kilometers, while the orbital height of the satellite is usually within the range of 160 kilometers to 2000 kilometers, much higher than the cloud height (heightc << heights). Therefore, the coverage area of the clouds can be approximately regarded as the surface area that needs to be reshot, and the cloud height (heightc) is usually assumed to be 10 kilometers.

[0124] Step S13-3, determine the area to be photographed according to the minimum shooting angle.

[0125] In this embodiment, the area to be photographed in the thick cloud coverage area is determined according to the minimum shooting angle.

[0126] Optionally, in combination with the above embodiment, an embodiment of the present application also provides another anti-cloud interference target detection method based on multi-satellite on-orbit scheduling. In this method, the "determine the area to be photographed according to the minimum shooting angle" in step S13-3 specifically includes steps S13-3-1 to step S13-3-2. In addition, this method further includes steps S13-4 and step S13-5:

[0127] Step S13-3-1, when the minimum shooting angle is not greater than the maximum side-sway angle of the satellite, determine the thick cloud coverage area as the area to be photographed, and the minimum shooting angle of the area to be photographed is the minimum shooting angle of the thick cloud coverage area.

[0128] In this embodiment, considering that the camera of the satellite is usually restricted by the maximum side-sway angle (for example, 30°) during the actual shooting process, when the shooting angle of the satellite exceeds the limit of the maximum side-sway angle, the captured image may have a large distortion. Therefore, in order to ensure the accuracy of the shooting effect and the image quality, it is necessary to add a limit to the width (L) of the area to be photographed in the thick cloud coverage area.

[0129] Figure 6This is a schematic diagram showing the maximum lateral sway angle as illustrated in this application. For details, please refer to [reference needed]. Figure 6 When the minimum shooting angle of the thick cloud-covered area is not greater than the maximum side angle of the satellite, it means that when shooting from the minimum shooting angle to the maximum side angle of the satellite, the resulting image will not be distorted. Therefore, there is no need to segment the thick cloud-covered area, and the thick cloud-covered area is the area to be shot.

[0130] Step S13-3-2: When the minimum shooting angle is greater than the maximum side angle of the satellite, the thick cloud-covered area is divided into multiple shooting areas with the aim that the minimum shooting angle of the area to be shot is not greater than the maximum side angle of the satellite. The minimum shooting angle of the area to be shot is determined based on the width of the area to be shot and the height of the cloud layer.

[0131] Specifically, when the minimum shooting angle of the thick cloud-covered area is greater than the satellite's maximum side-swing angle, it means that even if the shooting angle is set to the minimum, the resulting image will be distorted. Therefore, it is necessary to limit the width of the thick cloud-covered area and segment it. During segmentation, the goal is to ensure that the minimum shooting angle of the final image-to-be-shot area does not exceed the satellite's maximum side-swing angle. The thick cloud-covered area is then segmented into multiple image-to-be-shot areas. During this segmentation, the central satellite divides the thick cloud-covered area into several small rectangular regions (i.e., image-to-be-shot areas) based on its size, ensuring that the shooting angle of each small rectangular region is within the allowable range (i.e., the minimum shooting angle does not exceed the maximum side-swing angle). This avoids image distortion caused by excessively large shooting angles. Ultimately, the segmented thick cloud-covered area contains multiple image-to-be-shot areas. Similarly, the minimum shooting angle for each image-to-be-shot area is determined based on the width of the image-to-be-shot area and the cloud height.

[0132] The method further includes:

[0133] Step S13-4: When the thick cloud-covered area is the area to be photographed, determine an edge satellite as the target edge satellite, and match an edge satellite to the area to be photographed.

[0134] In this embodiment, when the area covered by thick clouds does not need to be segmented, the target edge satellite is a single edge satellite, indicating that only one edge satellite is needed to complete the imaging of the area covered by thick clouds.

[0135] Step S13-5: When the thick cloud-covered area is divided into multiple areas to be photographed, multiple edge satellites are identified as target edge satellites, and multiple areas to be photographed are matched with multiple edge satellites respectively.

[0136] Figure 7This is a schematic diagram illustrating the segmentation of a thick cloud-covered area, as shown in this application. Figure 7 In this embodiment, when a thick cloud-covered area needs to be segmented, the number of target edge satellites is equal to the number of areas to be photographed, and each target edge satellite performs photography of one of the areas to be photographed. Specifically, in Figure 7 In the image, solid lines represent the thick cloud-covered area to be photographed, while dashed lines simulate the process of dividing the thick cloud-covered area into multiple smaller patches (i.e., multiple areas to be photographed). Each patch is photographed by a different edge satellite. Figure 7 The image only shows a schematic diagram of two small sections.

[0137] In summary, by combining oblique-view imagery and regional segmentation scheduling strategies, this application effectively overcomes the cloud cover problem, improves the surface coverage capability of remote sensing satellites under complex meteorological conditions, and ensures image quality and observation accuracy. This innovative method has broad application prospects in cloud monitoring, disaster early warning, and meteorological observation.

[0138] Optionally, in conjunction with the above embodiments, one embodiment of this application also provides another target detection method based on multi-satellite on-orbit scheduling to resist cloud interference. In this method, before performing step S14, "generating a shooting schedule for the area to be photographed", it further includes steps S21-S25:

[0139] Step S21: For each of the multiple edge satellites, obtain the shooting angle change information and time interval of the edge satellite; the shooting angle change information includes: the shooting angle of the edge satellite during the last shooting and the minimum shooting angle of the area to be shot; the time interval is: the time required for the edge satellite to turn from the current shooting angle to the maximum side angle of the satellite, and the current shooting angle is the angle of the camera of the edge satellite when the area to be shot is determined.

[0140] As mentioned earlier, by limiting the minimum shooting angle and the maximum side-swing angle, it is determined whether the area covered by thick clouds needs to be segmented, thus identifying the area to be photographed. Therefore, after obtaining the area to be photographed, the shooting angle of the edge satellites performing the photographing process for that area is limited to between the minimum shooting angle and the maximum side-swing angle. In this embodiment, to further optimize the scheduling of edge satellites and ensure that each area to be photographed can select the optimal edge satellite as the target edge satellite for photographing, the selection process of the target edge satellite is further restricted by introducing shooting angle change information to obtain the optimal edge satellite scheduling scheme.

[0141] Specifically, it is necessary to obtain the shooting angle of each edge satellite during its last shooting, the minimum shooting angle of the area to be photographed, and the time interval as the shooting angle change information of each edge satellite. In particular, to determine the time interval, after determining the area to be photographed, it is necessary to obtain the current shooting angle of each edge satellite, and then, based on the current shooting angle of each edge satellite and the maximum side swing angle of the satellite, calculate the time required for the shooting angle of each edge satellite's camera to turn from the current shooting angle to the maximum side swing angle of the satellite, and use this time as the time interval of that edge satellite.

[0142] Step S22: Determine the shooting angle torsion speed of the edge satellite based on the shooting angle change information and the time interval.

[0143] In this embodiment, based on the angle difference and time interval in the shooting angle change information, the shooting angle torsion speed required for the cluster edge satellite to reach the minimum shooting angle of the area to be shot from the shooting angle when it completed the last shooting can be determined.

[0144] Step S23: Select multiple edge satellites whose shooting angle twist speed is not greater than the maximum speed of the satellite's camera angle twist as initial edge satellites.

[0145] In this embodiment, after determining the shooting angle torsion speed of each satellite relative to the area to be photographed, for each area to be photographed, multiple edge satellites whose shooting angle torsion speed is not greater than the maximum speed of the satellite's camera angle torsion are selected as initial edge satellites.

[0146] Step S24: Determine the target edge satellite and the target observation window from the initially selected edge satellites.

[0147] In this embodiment, for each area to be photographed, one edge satellite is selected from the initial edge satellites as the target edge satellite for that area. The target observation window of the target edge satellite is from the minimum shooting angle of the area to be photographed to the maximum side angle of the satellite.

[0148] Step S25: Determine the target acquisition time based on the current position of the target edge satellite.

[0149] In this embodiment, after determining the target imaging satellite and the target observation window, it is also necessary to determine the target imaging time for the target imaging satellite to photograph the area to be photographed. The target imaging time is determined based on the current position of the target edge satellite. Based on the current position, it is determined how long it takes for the target edge satellite to move from its current position to a position where it can photograph the area to be photographed with the maximum side swing angle. Combined with the current time, the target imaging time for the target edge satellite to perform the photographing is calculated.

[0150] The target observation window is determined based on the available observation window of the target edge satellite; the available observation window of the target edge satellite represents the range of the target edge satellite's shooting angle from the minimum shooting angle of the area to be photographed to the maximum side sway angle of the satellite.

[0151] Optionally, in conjunction with the above embodiments, one embodiment of this application also provides another cloud-interference-resistant target detection method based on multi-satellite on-orbit scheduling. In this method, the process of selecting multiple edge satellites as initial edge satellites based on the shooting angle torsion speed not exceeding the maximum speed of the satellite's camera angle torsion is as follows:

[0152] ;

[0153] in, This indicates the shooting angle of the last time the edge satellite was photographed. The minimum shooting angle of the area to be photographed; Indicates the time interval; This indicates the maximum speed at which the camera angle of the satellite twists.

[0154] Among them, edge satellites that satisfy the above inequality are identified as preliminary edge satellites.

[0155] Optionally, in conjunction with the above embodiments, one embodiment of this application also provides another cloud-interference-resistant target detection method based on multi-satellite on-orbit scheduling. In this method, the step S24 of "determining the target edge satellite from the initially selected edge satellites" specifically includes steps S24-1 to S24-2:

[0156] Step S24-1: When there are multiple preliminary edge satellites, obtain the available observation window of the preliminary edge satellite; the available observation window of the preliminary edge satellite represents the range from the first shooting angle of the preliminary edge satellite to the maximum side swing angle of the satellite when shooting the area to be shot.

[0157] In this embodiment, when there is only one initially selected edge satellite, it is directly used as the target edge satellite for the area to be photographed, without further screening. When there are multiple initially selected edge satellites, further screening is required to obtain the most suitable initially selected edge satellite for photographing the area to be photographed as the target edge satellite.

[0158] Figure 8 This is a schematic diagram illustrating the available observation windows shown in this application, such as... Figure 8As shown, specifically, when there are multiple initial edge satellites, it is necessary to obtain the available observation window for each initial edge satellite. The available observation window represents the range from the first shooting angle of the initial edge satellite to the maximum side angle of the satellite when it takes pictures of the area to be photographed. The first shooting angle is the minimum shooting angle required for the initial edge satellite to take pictures of the area to be photographed.

[0159] It should be noted that the first shooting angle is not the same as the minimum shooting angle of the area to be photographed. When the initial edge satellite and the central satellite are in the same orbit, the first shooting angle is equal to the minimum shooting angle of the area to be photographed. When the initial edge satellite and the central satellite are not in the same orbit, the initial edge satellite and the area to be photographed are not parallel. Therefore, when the initial edge satellite needs to photograph the entire area to be photographed, the "width" of the area to be photographed that the initial edge satellite needs to cover is actually longer than the width (L) of the area to be photographed. The corresponding first shooting angle will be greater than the minimum shooting angle of the area to be photographed. Therefore, the available observation window of the initial edge satellite is the range from the first shooting angle of the initial edge satellite to the maximum side angle of the satellite. The range of the available observation window is no greater than the range from the minimum shooting angle to the maximum side angle of the area to be photographed.

[0160] Step S24-2: Select the final edge satellites. The final edge satellites are the initial edge satellites with the largest available observation window. The final edge satellites are used as the target edge satellites.

[0161] In this embodiment, in some cases, to improve imaging efficiency and cover a wider area of ​​the Earth's surface, the target edge satellite needs to have a larger available observation window. The available observation window represents the range from the edge satellite's yaw angle (i.e., the first imaging angle) when it can completely capture the area to be photographed to the edge satellite's maximum yaw angle. To ensure longer imaging time and better imaging conditions, the central satellite will prioritize allocating imaging tasks to the initially selected edge satellites with larger available observation windows during the scheduling process.

[0162] Therefore, after obtaining the available observation windows of each preliminary edge satellite, the preliminary edge satellite with the largest available observation window can be selected as the final edge satellite, and the final edge satellite can be used as the target edge satellite.

[0163] For example, the process of selecting the initial edge satellite with the largest available observation window is as follows: ,in, Indicates the maximum yaw angle. This indicates the first shooting angle.

[0164] Optionally, in conjunction with the above embodiments, one embodiment of this application also provides another cloud-interference-resistant target detection method based on multi-satellite on-orbit scheduling. In this method, after "screening out the final edge satellites" in step S24-2, steps S31-S32 are further included:

[0165] Step S31: When there are multiple final edge satellites, for each final edge satellite, obtain the shooting angle of the last shooting of the final edge satellite and the minimum shooting angle of the area to be shot.

[0166] In this embodiment, when there is only one final edge satellite, it is directly used as the target edge satellite for the area to be photographed, without further selection. When there are multiple final edge satellites, further selection is required to obtain the most suitable final edge satellite for photographing the area to be photographed as the target edge satellite.

[0167] Specifically, when multiple final edge satellites are involved, it is necessary to obtain the shooting angle of the last shooting of each final edge satellite and the minimum shooting angle of the area to be photographed.

[0168] Step S32: Select the final edge satellite with the smallest camera twist angle between the shooting angle of the last shooting of the final edge satellite and the minimum shooting angle of the area to be shot as the target edge satellite.

[0169] Before capturing images, edge satellites need to adjust their shooting angle from the previous angle to the minimum shooting angle of the target area. This angle change control strategy is particularly important because satellite camera rotation is typically limited by its mechanical structure and orbital dynamics. Rapid and frequent angle adjustments can lead to image blurring, unstable field of view, and even adversely affect target detection accuracy. Therefore, the angle change process increases the mechanical adjustment burden on the edge satellite's camera. It is necessary to control the magnitude of the shooting angle change and reduce the drastic changes between shooting missions to improve shooting accuracy and reduce potential errors caused by rapid rotation.

[0170] Based on this, during the scheduling process, the central satellite will prioritize shooting schemes that can minimize changes in camera torsional angle, thereby achieving smoother and more accurate target shooting and image capture.

[0171] When there are multiple final edge satellites, the shooting angle change value between the shooting angle of the last shooting of each final edge satellite and the minimum shooting angle of the area to be shot is determined by obtaining the shooting angle of the last shooting of each final edge satellite and the minimum shooting angle of the area to be shot.

[0172] Finally, the edge satellite with the smallest change in shooting angle between the shooting angle at the last shooting angle and the minimum shooting angle of the area to be shot is selected as the target edge satellite.

[0173] For example, the process of selecting the final edge satellite with the smallest change in shooting angle as the target edge satellite is as follows: ,in, This indicates the shooting angle of the last image taken by the edge satellite in the final selection process. The minimum shooting angle for the area to be photographed.

[0174] Optionally, in conjunction with the above embodiments, one embodiment of this application also provides another target detection method based on multi-satellite on-orbit scheduling to resist cloud interference. In this method, the step S16 above, "performing target detection on the thin cloud region image and the target image to determine the target in the region to be detected," specifically includes steps S16-1 to S16-4:

[0175] Step S16-1: Perform cloud removal processing on the thin cloud area image using the central satellite.

[0176] Specifically, the working process of the central satellite is divided into two stages: the first stage is the processing of images of thin cloud areas, and the second stage is the processing of images of thick cloud areas.

[0177] Figure 9 This is a schematic diagram illustrating the first stage of cloud removal and target recognition processes shown in this application, as follows: Figure 9 As shown, in this embodiment, for thin cloud areas, the central satellite uses a cloud removal algorithm to remove or reduce the impact of clouds on the image in the thin cloud area.

[0178] Step S16-2: Target detection is performed on the thin cloud region image after cloud removal processing to determine the target in the thin cloud region image after cloud removal processing. The image region where the target is located in the thin cloud region image is segmented to obtain the first target region image.

[0179] In this embodiment, the central satellite performs target detection on the thin cloud region image after cloud removal processing. During this process, the YOLO target detection algorithm can be used to analyze and extract the target in the thin cloud region image, and the image region where the target is located in the thin cloud region image is segmented from the thin cloud region image to obtain the first target region image, which represents the image region where the target is located in the thin cloud region image.

[0180] Step S16-3: Target detection is performed on the target image to determine the target in the target image. The image region where the target is located in the target image is segmented to obtain a second target region image.

[0181] Figure 10 This is a schematic diagram illustrating the second stage of the target identification process shown in this application, as follows: Figure 10 As shown, in this embodiment, the target edge satellite performs target detection on the target image. In this process, the YOLO target detection algorithm can also be used to analyze and extract the target in the target image, and the image region where the target is located in the target image is segmented from the target image to obtain a second target region image. The second target region image represents the image region where the target is located in the target image.

[0182] Step S16-4, the method further includes: sending the first target area image and the second target area image to the ground station through multiple transmission queues respectively.

[0183] In this embodiment, during the process of establishing a connection between the central satellite and the ground station, the central satellite sends the first target area image to the ground station through the transmission queue corresponding to the central satellite. Similarly, during the process of establishing a connection between the target edge satellite and the ground station, the target edge satellite sends the second target area image to the ground station through the transmission queue corresponding to the target edge satellite, so that the first target area image and the second target area image in the transmission queue are transmitted back to the ground station for further analysis and processing.

[0184] The process of determining the target edge satellites mentioned above is summarized below. Figure 11 This is a schematic diagram illustrating the priority relationship of multi-level constraints as shown in this application, such as... Figure 11 As shown, in the process of selecting target edge satellites, firstly, constraint 1 limits the width of the thick cloud coverage area so that the minimum shooting angle is not greater than the maximum side angle of the satellite. At the same time, constraint 2 also needs to be satisfied, which means that the shooting angle torsion speed of the selected target edge satellite is not greater than the maximum speed of the satellite's camera angle torsion.

[0185] If multiple candidate edge satellites (i.e., multiple preliminary edge satellites) exist, and constraint 3 is applied to further filter them, the edge satellite with the largest available observation window is selected as the target edge satellite (i.e., the final edge satellite mentioned above) if constraints 1 and 2 are satisfied. If multiple candidate edge satellites still exist (i.e., multiple final edge satellites), the final edge satellite is selected as the target edge satellite if constraint 4 is applied to further filter them, the final edge satellite with the smallest change in shooting angle between the shooting angle at the time of the last shooting and the minimum shooting angle of the area to be photographed is selected.

[0186] Based on the same inventive concept, an embodiment of this application also provides a cloud-interference-resistant target detection system based on multi-satellite on-orbit scheduling. The system includes a satellite constellation architecture and a ground station. The satellite constellation architecture includes a central satellite and multiple edge satellites.

[0187] The central satellite is used to take pictures of the area to be detected, obtain initial remote sensing images, and identify and classify the cloud thickness in the initial remote sensing images to determine the thin cloud area images and thick cloud area images in the initial remote sensing images.

[0188] The central satellite is used to determine at least one area to be photographed from the thick cloud-covered area corresponding to the thick cloud area image;

[0189] The central satellite is used to generate a shooting schedule for the area to be photographed. The shooting schedule is used to indicate the target edge satellite, the target observation window, and the target shooting time for photographing the area to be photographed. The target edge satellite is one of the multiple edge satellites.

[0190] The target edge satellite is used to take pictures of the area to be photographed according to the shooting schedule to obtain a target image of the area to be photographed.

[0191] The central satellite is used to perform target detection on the thin cloud region image and determine the target in the region to be detected;

[0192] The target edge satellite is used to perform target detection on the target image and determine the target in the area to be detected.

[0193] Optionally, the central satellite is used to determine at least one area to be photographed from the thick cloud-covered area corresponding to the thick cloud area image, specifically including:

[0194] The central satellite is used to determine the thick cloud-covered area based on the latitude and longitude information of the thick cloud-covered area in the initial remote sensing image, and the thick cloud-covered area is a rectangular area of ​​latitude and longitude of the thick cloud-obscured area in the thick cloud-covered area image.

[0195] The central satellite is used to determine the minimum shooting angle of the thick cloud-covered area based on the width of the thick cloud-covered area and the height of the cloud layer.

[0196] The central satellite is used to determine the area to be photographed based on the minimum shooting angle.

[0197] Optionally, the central satellite is used to determine the area to be photographed based on the minimum shooting angle, specifically including:

[0198] The central satellite is used to determine the thick cloud-covered area as the area to be photographed when the minimum shooting angle is not greater than the maximum side angle of the satellite, and the minimum shooting angle of the area to be photographed is the minimum shooting angle of the thick cloud-covered area.

[0199] The central satellite is used to divide the thick cloud-covered area into multiple areas to be photographed, with the aim that the minimum shooting angle of the area to be photographed is not greater than the maximum side angle of the satellite when the minimum shooting angle is greater than the maximum side angle of the satellite. The minimum shooting angle of the area to be photographed is determined based on the width of the area to be photographed and the height of the cloud layer.

[0200] The central satellite is also used for:

[0201] When the thick cloud-covered area is the area to be photographed, an edge satellite is identified as the target edge satellite, and the area to be photographed is matched with an edge satellite;

[0202] When the thick cloud-covered area is divided into multiple areas to be photographed, multiple edge satellites are identified as target edge satellites, and multiple areas to be photographed are matched with multiple edge satellites respectively.

[0203] Optionally, the central satellite is also used for:

[0204] Before generating a shooting schedule for the area to be photographed, for each of the multiple edge satellites, the shooting angle change information and time interval of the edge satellite are obtained; the shooting angle change information includes: the shooting angle of the edge satellite during the last shooting and the minimum shooting angle of the area to be photographed; the time interval is: the time required for the edge satellite to turn from the current shooting angle to the maximum side angle of the satellite, and the current shooting angle is the angle of the camera of the edge satellite when the area to be photographed is determined;

[0205] The shooting angle torsion speed of the edge satellite is determined based on the shooting angle change information and the time interval.

[0206] Multiple edge satellites whose shooting angle torsion speed is not greater than the maximum speed of the satellite's camera angle torsion are selected as initial edge satellites;

[0207] The target edge satellite and the target observation window are determined from the initially selected edge satellites;

[0208] The target acquisition time is determined based on the current position of the target edge satellite;

[0209] The target observation window is determined based on the available observation window of the target edge satellite; the available observation window of the target edge satellite represents the range of the target edge satellite's shooting angle from the minimum shooting angle of the area to be photographed to the maximum side sway angle of the satellite.

[0210] Optionally, determining the target edge satellite from the initially selected edge satellites includes:

[0211] The central satellite is used to obtain the available observation window of the initial edge satellite when there are multiple initial edge satellites; the available observation window of the initial edge satellite represents the range from the first shooting angle of the initial edge satellite to the maximum side angle of the satellite when the initial edge satellite shoots the area to be shot.

[0212] The central satellite is used to select the final edge satellites, which are the initial edge satellites with the largest available observation window. The final edge satellites are then used as the target edge satellites.

[0213] Optionally, the central satellite is also used for:

[0214] After the final edge satellites are selected, when there are multiple final edge satellites, for each final edge satellite, obtain the shooting angle of the last shooting of the final edge satellite and the minimum shooting angle of the area to be shot;

[0215] The final edge satellite with the smallest camera twist angle between the shooting angle of the last shooting of the final edge satellite and the minimum shooting angle of the area to be photographed is selected as the target edge satellite.

[0216] Optionally, the plurality of edge satellites includes a first edge satellite group and a second edge satellite group; each edge satellite in the first edge satellite group operates along the same orbit as the central satellite, and each edge satellite in the second edge satellite group operates along an orbit with an inclination angle to the orbit of the central satellite;

[0217] The satellite constellation architecture is a satellite array centered on the central satellite, formed by a first edge satellite group and a second edge satellite group. The satellite constellation architecture is used to observe areas covered by thick clouds from multiple angles.

[0218] Optionally, the central satellite is used to perform target detection on the thin cloud region image to determine the target in the region to be detected. Specifically, the central satellite is used to perform cloud removal processing on the thin cloud region image; perform target detection on the cloud-removed thin cloud region image to determine the target in the cloud-removed thin cloud region image; segment the image region where the target is located in the thin cloud region image to obtain a first target region image; the central satellite is also used to send the first target region image to the ground station through a transmission queue.

[0219] Optionally, the target edge satellite is used to perform target detection on the target image and determine the target in the region to be detected. Specifically, this includes: performing target detection on the target image, determining the target in the target image, and segmenting the image region where the target is located in the target image to obtain a second target region image; the target edge satellite is also used to: transmit the first target region image and the second target region image to the ground station through multiple transmission queues respectively.

[0220] As the system is basically similar to the method embodiment, it is described in a relatively simple way. For relevant details, please refer to the description of the method embodiment.

[0221] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0222] Those skilled in the art will understand that embodiments of this application can be provided as methods, apparatus, or computer program products. Therefore, embodiments of this application can take the form of entirely hardware embodiments, entirely software embodiments, or embodiments combining software and hardware aspects. Furthermore, embodiments of this application can take the form of computer program products implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0223] This application describes embodiments with reference to flowchart illustrations and / or block diagrams of methods, terminal devices (systems), and computer program products according to embodiments of this application. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing terminal device to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing terminal device, generate instructions for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0224] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing terminal device to operate in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0225] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal equipment, causing a series of operational steps to be performed on the computer or other programmable terminal equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable terminal equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0226] Although preferred embodiments of the present application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the embodiments of the present application.

[0227] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or terminal device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or terminal device. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or terminal device that includes the element.

[0228] The above provides a detailed description of the cloud-interference-resistant target detection method and system based on multi-satellite on-orbit scheduling provided in this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the method and its core ideas. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A cloud layer interference resistant target detection method based on multi-satellite on-orbit scheduling, characterized in that, The method includes: Deploy a satellite constellation architecture, which includes a central satellite and multiple edge satellites; The central satellite takes pictures of the area to be detected to obtain initial remote sensing images, and identifies and classifies the cloud thickness in the initial remote sensing images to determine the thin cloud area images and thick cloud area images in the initial remote sensing images. Using the central satellite, at least one area to be photographed is determined from the thick cloud-covered area corresponding to the thick cloud area image; Using the central satellite, a shooting schedule is generated for the area to be photographed. The shooting schedule is used to indicate the target edge satellite, the target observation window, and the target shooting time for photographing the area to be photographed. The target edge satellite is one of the multiple edge satellites. According to the shooting schedule, the target area is photographed by the target edge satellite to obtain the target image of the target area; Target detection is performed on the thin cloud region image and the target image to determine the target in the region to be detected. 2.The cloud layer interference resistant target detection method based on multi-satellite on-orbit scheduling according to claim 1, wherein, Determining at least one area to be photographed from the thick cloud coverage area corresponding to the thick cloud area image includes: For the image of the thick cloud region, the central satellite determines the thick cloud-covered area based on the latitude and longitude information of the thick cloud region image in the initial remote sensing image. The thick cloud-covered area is a rectangular area of ​​latitude and longitude of the thick cloud-obscured area in the image of the thick cloud region. The minimum shooting angle for the thick cloud-covered area is determined based on the width of the thick cloud-covered area and the height of the cloud layer. The area to be photographed is determined based on the minimum shooting angle. 3.The cloud layer interference resistant target detection method based on multi-satellite on-orbit scheduling according to claim 2, wherein, Determining the area to be photographed based on the minimum shooting angle includes: When the minimum shooting angle is not greater than the maximum side angle of the satellite, the thick cloud-covered area is determined as the area to be shot, and the minimum shooting angle of the area to be shot is the minimum shooting angle of the thick cloud-covered area. When the minimum shooting angle is greater than the maximum side angle of the satellite, the thick cloud-covered area is divided into multiple shooting areas with the aim that the minimum shooting angle of the area to be shot is not greater than the maximum side angle of the satellite. The minimum shooting angle of the area to be shot is determined based on the width of the area to be shot and the height of the cloud layer. The method further includes: When the thick cloud-covered area is the area to be photographed, an edge satellite is identified as the target edge satellite, and the area to be photographed is matched with an edge satellite; When the thick cloud-covered area is divided into multiple areas to be photographed, multiple edge satellites are identified as target edge satellites, and multiple areas to be photographed are matched with multiple edge satellites respectively.

4. The method of claim 3, wherein, Before generating a shooting schedule for the area to be shot, the process also includes: For each of the multiple edge satellites, obtain the shooting angle change information and time interval of the edge satellite; the shooting angle change information includes: the shooting angle of the edge satellite at the last shooting and the minimum shooting angle of the area to be shot; the time interval is: the time required for the edge satellite to turn from the current shooting angle to the maximum side angle of the satellite, and the current shooting angle is the angle of the camera of the edge satellite when the area to be shot is determined; The shooting angle torsion speed of the edge satellite is determined based on the shooting angle change information and the time interval. Multiple edge satellites whose shooting angle torsion speed is not greater than the maximum speed of the satellite's camera angle torsion are selected as initial edge satellites; The target edge satellite and the target observation window are determined from the initially selected edge satellites; The target acquisition time is determined based on the current position of the target edge satellite; The target observation window is determined based on the available observation window of the target edge satellite; the available observation window of the target edge satellite represents the range of the target edge satellite's shooting angle from the minimum shooting angle of the area to be photographed to the maximum side sway angle of the satellite.

5. The method of claim 4, wherein, The process of selecting multiple edge satellites whose shooting angle torsion speed is not greater than the maximum speed of the satellite's camera angle torsion as initial edge satellites is as follows: ； in, This indicates the shooting angle of the last time the edge satellite was photographed. The minimum shooting angle of the area to be photographed; Indicates the time interval; This indicates the maximum speed at which the camera angle of the satellite twists.

6. The method of claim 4, wherein, The step of determining the target edge satellite from the initially selected edge satellites includes: When there are multiple initial edge satellites, the available observation window of the initial edge satellite is obtained; the available observation window of the initial edge satellite represents the range from the first shooting angle of the initial edge satellite to the maximum side swing angle of the satellite when shooting the area to be shot. Select the final edge satellites, which are the initial edge satellites with the largest available observation window, and use the final edge satellites as the target edge satellites; The first shooting angle is the side angle at which the edge satellite can completely capture the area to be shot.

7. The method of claim 6, wherein the cloud layer interference resistant target detection method based on multi-satellite on-orbit scheduling is characterized in that, After the final selection of edge satellites, the process also includes: When there are multiple final edge satellites, for each final edge satellite, obtain the shooting angle of the last shooting of the final edge satellite and the minimum shooting angle of the area to be shot; The final edge satellite with the smallest change in shooting angle between the shooting angle of the last shooting of the final edge satellite and the minimum shooting angle of the area to be photographed is selected as the target edge satellite.

8. The cloud layer interference resistant target detection method based on multi-satellite on-orbit scheduling according to any one of claims 1-7, characterized in that, The plurality of edge satellites includes a first edge satellite group and a second edge satellite group; each edge satellite in the first edge satellite group operates along the same orbit as the central satellite, and each edge satellite in the second edge satellite group operates along an orbit with an inclination angle to the orbit of the central satellite; The satellite constellation architecture is a satellite array centered on the central satellite, formed by a first edge satellite group and a second edge satellite group. The satellite constellation architecture is used to observe areas covered by thick clouds from multiple angles.

9. The cloud layer interference resistant target detection method based on multi-satellite on-orbit scheduling according to any one of claims 1-7, characterized in that, The step of performing target detection on the thin cloud region image and the target image to determine the target in the region to be detected includes: The image of the thin cloud area is processed by the central satellite to remove clouds. Target detection is performed on the thin cloud region image after cloud removal processing to determine the target in the thin cloud region image. The image region where the target is located in the thin cloud region image is segmented to obtain the first target region image. Target detection is performed on the target image to identify the target in the target image, and the image region where the target is located in the target image is segmented to obtain a second target region image; The method further includes: sending the first target area image and the second target area image to the ground station through multiple transmission queues respectively. 10.A cloud layer interference resistant target detection system based on multi-satellite on-orbit scheduling, characterized in that, The system includes a satellite constellation architecture and a ground station, wherein the satellite constellation architecture comprises a central satellite and multiple edge satellites; The central satellite is used to take pictures of the area to be detected, obtain initial remote sensing images, and identify and classify the cloud thickness in the initial remote sensing images to determine the thin cloud area images and thick cloud area images in the initial remote sensing images. The central satellite is used to determine at least one area to be photographed from the thick cloud-covered area corresponding to the thick cloud area image; The central satellite is used to generate a shooting schedule for the area to be photographed. The shooting schedule is used to indicate the target edge satellite, the target observation window, and the target shooting time for photographing the area to be photographed. The target edge satellite is one of the multiple edge satellites. The target edge satellite is used to take pictures of the area to be photographed according to the shooting schedule to obtain a target image of the area to be photographed. The central satellite is used to perform target detection on the thin cloud region image and determine the target in the region to be detected; The target edge satellite is used to perform target detection on the target image and determine the target in the area to be detected.

Images