A fast detection method for large-area marine fixed targets based on multi-source satellite data

Abstract

An embodiment of the present invention discloses a method for rapidly detecting fixed targets at sea over a large area based on multi-source satellite data. The method comprises: acquiring radar satellite data from multiple periods of the target sea area; performing threshold segmentation on the radar satellite data from each period to distinguish the sea surface area and targets at sea from each period; performing spatial overlay analysis on the targets at sea from each period to preliminarily identify fixed targets; utilizing AIS data from ship satellites from the same period to eliminate mobile ships from the preliminarily identified fixed targets to obtain the final fixed targets; extracting the region containing the final fixed target from the optical satellite data of the target sea area; and inputting the extracted region data into a deep learning image detection model to detect the type of the fixed target. This embodiment reduces the amount of data processing required for large sea areas.

Description

Technical Field

[0001] The embodiments of the present invention relate to the field of target positioning technology, and in particular to a method for rapidly detecting fixed targets at sea in a large area based on multi-source satellite data. Background Art

[0002] Fixed targets at sea refer to objects that are exposed on the sea surface and have a fixed position, such as offshore drilling platforms, lighthouses, and fish farms. Fixed target detection at sea involves detecting the presence of fixed targets on the sea surface and identifying their type, such as an offshore drilling platform, lighthouse, or fish farm.

[0003] In the existing technology, optical remote sensing images of the sea area are acquired by optical satellites, and an automatic detection method based on machine learning is used to detect marine targets in the optical remote sensing images.

[0004] This method has good effects in identifying targets at sea in small areas. However, when it is applied to large sea areas, a large amount of optical satellite data needs to be processed by orthorectification and other methods. However, the actual useful data after processing is very small, which is time-consuming and labor-intensive. In addition, the identification of targets is very blind and prone to misjudgment. Summary of the Invention

[0005] An embodiment of the present invention provides a method for rapid detection of fixed targets at sea in a large area based on multi-source satellite data. The method filters optical satellite data by pre-identifying fixed targets, reduces the workload of optical image data processing, and rapidly detects targets.

[0006] In a first aspect, an embodiment of the present invention provides a method for rapidly detecting fixed targets at sea in a large area based on multi-source satellite data, comprising:

[0007] Acquire multiple phases of radar satellite data of the target sea area;

[0008] Threshold segmentation is performed on radar satellite data of each period to distinguish the sea surface area and sea targets of each period;

[0009] Conduct spatial superposition analysis on maritime targets of each phase and preliminarily identify fixed targets among the maritime targets;

[0010] Using the satellite AIS data of ships in the same period, the moving ships are eliminated from the initially identified fixed targets to obtain the final fixed targets;

[0011] Extracting the area where the final fixed target is located from the optical satellite data of the target sea area;

[0012] The extracted area data is input into a deep learning image detection model to detect the type of the fixed target.

[0013] In a second aspect, an embodiment of the present invention provides an electronic device, including:

[0014] one or more processors;

[0015] a memory for storing one or more programs,

[0016] When the one or more programs are executed by the one or more processors, the one or more processors implement the above-mentioned large-area maritime fixed target rapid detection method based on multi-source satellite data.

[0017] In a third aspect, an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-mentioned method for rapid detection of large-area maritime fixed targets based on multi-source satellite data.

[0018] The embodiment of the present invention combines the advantages of radar, ship AIS and optical multi-source satellite data, screens fixed targets at sea layer by layer, gradually improves the positioning accuracy of fixed targets, and finally detects the specific type of fixed targets. Specifically, first, the huge difference between the background and the target in the radar satellite data is used to quickly circle the location of the target object; then, spatial overlay analysis is used to determine the fixed target among the targets; then, the ship-borne AIS equipment is used to accurately locate the ship, effectively eliminate the interference information of ships at sea, and reduce radar information misjudgment; finally, deep learning technology is combined with optical imaging to quickly detect the category to which the target belongs. This method is particularly suitable for the rapid identification of fixed targets in large sea areas. It can extract valid data from a large amount of sea area data, greatly improving the efficiency and accuracy of fixed target detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the specific embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0020] Figure 1 The present invention provides a flowchart of a method for rapidly detecting fixed targets at sea in a large area based on multi-source satellite data.

[0021] Figure 2 It is a structural diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0022] To make the objectives, technical solutions, and advantages of the present invention more clear, the technical solutions of the present invention are described clearly and completely below. Obviously, the embodiments described are only some of the embodiments of the present invention, not all of them. All other embodiments derived by persons of ordinary skill in the art based on the embodiments of the present invention without inventive effort are also within the scope of protection of the present invention.

[0023] In the description of the present invention, it should be noted that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicating orientations or positional relationships, are based on the orientations or positional relationships shown in the accompanying drawings and are intended solely to facilitate and simplify the description of the present invention. They are not intended to indicate or imply that the devices or components referred to must have, be constructed, or operate in a specific orientation, and therefore should not be construed as limitations on the present invention. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0024] In the description of the present invention, it should also be noted that, unless otherwise expressly specified or limited, the terms "mounted," "connected," and "connected" should be understood broadly. For example, they may refer to fixed, detachable, or integral connections; mechanical or electrical connections; direct or indirect connections through an intermediate medium; and internal communication between two components. Those skilled in the art will understand the specific meanings of the above terms in the present invention based on the specific circumstances.

[0025] Figure 1 This is a flowchart of a method for rapidly detecting fixed targets at sea in a large area based on multi-source satellite data provided by an embodiment of the present invention. The method is applicable to detecting fixed targets at sea in a large area using optical satellite data. The method is executed by an electronic device, such as Figure 1 As shown, the specific steps include the following.

[0026] S110. Acquire multiple phases of radar satellite data of the target sea area.

[0027] The target sea area here refers to a large area of ​​sea, typically corresponding to multi-view radar satellite data, with each view covering a portion of the target sea area. Multiple periods refer to multiple time points or time periods, such as one month. This embodiment obtains multi-period, multi-view radar satellite data covering the target sea area as a data source for subsequent processing. For example, the publicly available Sentinel-1 satellite intensity map product data is used as the data source.

[0028] Optionally, after obtaining multi-period and multi-view radar satellite data, the multi-view radar satellite data of the same period are spliced ​​and mosaicked according to the geographical location to obtain the radar satellite data of the same period of the target sea area.

[0029] S120: performing threshold segmentation on the radar satellite data of each period to distinguish the sea surface area and sea targets of each period.

[0030] Radar satellite data is typically displayed as SAR grayscale images. The sea surface reflects radar signals weakly, appearing as a dark background in SAR grayscale images. However, targets at sea reflect radar signals strongly, appearing as bright spots. Based on this principle, this step uses threshold segmentation of the grayscale image to distinguish between the sea surface and targets at sea.

[0031] Optionally, the grayscale values ​​of the sea surface area in the radar satellite data are calibrated. Using the calibrated grayscale value as a threshold, Otsu threshold segmentation is performed on any period of radar satellite data to obtain a binary image of the sea surface area and maritime targets for that period. The Otsu threshold segmentation method is simple and fast, and can quickly extract maritime targets, both fixed and moving, that are exposed on the sea surface within a sea area.

[0032] S130: Perform spatial overlay analysis on the maritime targets of each period, and preliminarily identify fixed targets among the maritime targets.

[0033] Fixed targets are fixed and remain unchanged across different periods; mobile targets, on the other hand, are mobile and may correspond to different positions across different periods. Based on the above principles, this step uses spatial overlay analysis to perform the initial identification of fixed targets from the extracted maritime targets.

[0034] Optionally, spatial overlay analysis is performed using radar satellite data from each period as layers; if a maritime target appears in the layers of each period and their outlines overlap, the maritime target is identified as a fixed target; otherwise, the maritime target is identified as a moving target.

[0035] S140: Using the satellite AIS data of the ships in the same period, the mobile ships are eliminated from the initially identified fixed targets to obtain the final fixed targets.

[0036] Satellite AIS data is regularly uploaded by AIS equipment aboard ships, recording their movements, including their positions at various points in time. While most moving targets have been eliminated through spatial overlay analysis in S130, some moving vessels may still be present due to interference from factors like sea waves and timing. This step utilizes AIS data to eliminate these misidentified moving vessels and achieve secondary identification of fixed targets.

[0037] Optionally, the specific process of eliminating mobile ships includes: extracting the position of a preliminarily identified fixed target from any period of radar satellite data; searching for a ship position that coincides with the position from the ship satellite AIS data of the same period; if the time point corresponding to the ship position in the AIS data is the same as the time point of the fixed target in the radar satellite data, the preliminarily identified fixed target is re-identified as a mobile ship and eliminated.

[0038] S150: Extracting the area where the final fixed target is located from the optical satellite data of the target sea area.

[0039] Optical satellite data volumes are enormous for large ocean areas. The S110-S140 sensors eliminate interference from moving targets and accurately lock onto fixed targets. Extracting these areas from the massive optical data stream for subsequent processing reduces the processing scope from hundreds of scenes to just a few, significantly improving target detection speed and accuracy.

[0040] Optionally, first, an initial area covering the final fixed target is extracted from the optical satellite data of the target sea area. Exemplarily, the initial area can be data from certain satellites or certain scenes, with different scenes corresponding to different surface areas, which is related to the resolution of the specific satellite.

[0041] Then, orthophoto fusion processing is performed on the optical satellite data of the initial area to obtain an optical color image of the initial area. Specifically, the initial image of the optical satellite data is divided into panchromatic imagery and multispectral imagery. The two images have different resolutions and geometric position deviations. Through orthophoto fusion processing, position correction and resolution fusion are performed to obtain a standard image with true color and consistent resolution.

[0042] Finally, according to the type of fixed target of interest, the optical color image of the initial area is cropped with a maximum rectangular frame, and the cropped optical color image is used as the extracted regional data. Specifically, after obtaining each standard image, its spatial range is still very large. For example, the usual satellite width is 30km×30km, or 60km×60km, etc. In actual applications, the fixed targets of interest include offshore wind turbines, buoys, lighthouses, farms, offshore drilling platforms, etc., and their size does not exceed 500 meters. Most of the data in the standard image is still the sea surface background, and the fixed targets only occupy a very small area. Therefore, this step uses the target position obtained in S140 to crop the standard image with a maximum rectangular frame to further reduce the amount of data. Exemplarily, the size of the rectangular frame is selected to be 500 meters.

[0043] S160: Input the extracted area data into a deep learning image detection model to detect the type of the fixed target.

[0044] After obtaining the cropped optical image, the target type in the optical image is accurately detected based on the deep learning image detection model.

[0045] Alternatively, remote sensing imagery can be used to capture multiple images of fixed offshore targets (including offshore wind turbines, buoys, lighthouses, and fish farms) to form a sample set. The fixed target type of each sample is labeled. Each sample is then trained using a deep learning image detection model, ensuring that the model output consistently approximates the labeled fixed target type. In practice, the number of valuable fixed target types at sea is relatively limited, so the labeled sample set is also limited to ensure operability.

[0046] In a specific embodiment, the FasterR-CNN algorithm is trained using a sample set to obtain a final FasterR-CNN model; the trained FasterR-CNN model is used to detect the optical image obtained by S150, and finally the corresponding fixed target type at sea is output.

[0047] This embodiment combines the advantages of radar, ship AIS, and optical multi-source satellite data to screen fixed targets at sea layer by layer, gradually improve the positioning accuracy of fixed targets, and ultimately detect the specific type of fixed targets. Specifically, first, the huge difference between the background and the target in the radar satellite data is used to quickly circle the location of the target object; then, spatial overlay analysis is used to determine the fixed target among the targets; then, the ship-borne AIS equipment is used to accurately locate the ship, effectively eliminating interference information from ships at sea and reducing radar information misjudgment; finally, deep learning technology is combined with optical imaging to quickly detect the category to which the target belongs. This method is particularly suitable for the rapid identification of fixed targets in large sea areas. It can extract valid data from a large amount of sea area data, greatly improving the efficiency and accuracy of fixed target detection.

[0048] Figure 2 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention is shown in FIG. Figure 2 As shown, the device includes a processor 50, a memory 51, an input device 52 and an output device 53; the number of processors 50 in the device can be one or more. Figure 2 In the embodiment, a processor 50 is used as an example; the processor 50, the memory 51, the input device 52 and the output device 53 in the device can be connected by a bus or other means. Figure 2 The bus connection is taken as an example.

[0049] Memory 51, as a computer-readable storage medium, can be used to store software programs, computer-executable programs, and modules, such as the program instructions / modules corresponding to the method for rapid detection of large-area maritime stationary targets based on multi-source satellite data in the embodiments of the present invention. Processor 50 executes the software programs, instructions, and modules stored in memory 51 to perform various functional applications and data processing of the device, thereby implementing the aforementioned method for rapid detection of large-area maritime stationary targets based on multi-source satellite data.

[0050] The memory 51 may primarily include a program storage area and a data storage area. The program storage area may store an operating system and at least one application required for a function; the data storage area may store data generated based on the use of the terminal, etc. Furthermore, the memory 51 may include high-speed random access memory and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other non-volatile solid-state memory device. In some instances, the memory 51 may further include memory remotely located relative to the processor 50, and these remote memories may be connected to the device via a network. Examples of the aforementioned networks include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0051] The input device 52 may be used to receive input digital or character information and generate key signal input related to user settings and function control of the device. The output device 53 may include a display device such as a display screen.

[0052] An embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon. When the program is executed by a processor, the method for rapid detection of large-area marine fixed targets based on multi-source satellite data according to any embodiment is implemented.

[0053] The computer storage medium of the embodiment of the present invention can adopt any combination of one or more computer-readable media. Computer-readable media can be computer-readable signal media or computer-readable storage media. Computer-readable storage media can be, for example, but not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or components, or any combination thereof. More specific examples (non-exhaustive list) of computer-readable storage media include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In this document, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by an instruction execution system, device or device or used in combination with it.

[0054] A computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, which carries computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium that can transmit, propagate, or transport a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0055] The program code included in the computer-readable medium can be transmitted with any appropriate medium, including but not limited to wireless, electric wire, optical cable, RF, etc., or any suitable combination thereof. The computer program code for performing the operation of the present invention can be written in one or more programming languages ​​or a combination thereof, and the programming language includes an object-oriented programming language such as Java, Smalltalk, C++, and also includes a conventional procedural programming language such as "C" language or similar programming language. The program code can be executed entirely on the user's computer, partially on the user's computer, as an independent software package, partially on the user's computer and partially on a remote computer, or completely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computer (e.g., using an Internet service provider to connect via the Internet).

[0056] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the technical solutions of the embodiments of the present invention.

Claims

1. A method for rapid detection of fixed targets at sea in a large area based on multi-source satellite data, characterized in that: include: Acquire multiple phases of radar satellite data of the target sea area; Performing threshold segmentation on the radar satellite data of each period to distinguish the sea surface area and sea targets of each period; wherein the threshold is a grayscale threshold; Perform spatial overlay analysis on the maritime targets of each period to preliminarily identify fixed targets among the maritime targets. Specifically, the radar satellite data of each period is used as a layer for spatial overlay analysis. If a maritime target appears in all the layers of each period and the outlines overlap, the maritime target is identified as a fixed target; otherwise, the maritime target is identified as a moving target. Using the ship satellite AIS data from the same period, mobile ships are eliminated from the initially identified fixed targets to obtain the final fixed targets. Specifically, the position of the initially identified fixed target is extracted from any period of radar satellite data. The position of a ship that coincides with the position is searched from the ship satellite AIS data from the same period. If the time point corresponding to the ship position in the AIS data is the same as the time point in the radar satellite data, the initially identified fixed target is re-identified as a mobile ship. Extracting the area where the final fixed target is located from the optical satellite data of the target sea area; specifically, extracting an initial area covering the final fixed target from the optical satellite data of the target sea area; performing ortho fusion processing on the optical satellite data of the initial area to obtain an optical color image of the initial area; performing maximum rectangular frame cropping on the optical color image of the initial area according to the type of the fixed target of interest, and using the cropped optical color image as the extracted area data; The extracted area data is input into a deep learning image detection model to detect the type of the fixed target.

2. The method according to claim 1, characterized in that The target sea area corresponds to multi-view radar satellite data; The method of obtaining multiple phases of radar satellite data of the target sea area includes: Acquire multi-phase, multi-view radar satellite data covering the target sea area; The multi-view radar satellite data of the same period are spliced ​​and mosaicked according to the geographical location to obtain the radar satellite data of the same period of the target sea area.

3. The method according to claim 1, characterized in that The radar satellite data is presented as a grayscale image; The threshold segmentation of the radar satellite data of each period to distinguish the sea surface area and sea targets of each period includes: Calibrate the grayscale value of the sea surface area in radar satellite data; Using the calibrated grayscale value as the threshold, Otsu threshold segmentation is performed on the radar satellite data of any period to obtain the binary images of the sea surface area and the sea targets of the period.

4. The method according to claim 1, wherein The types of fixed targets of interest include at least one of the following: offshore wind turbines, buoys, lighthouses, fish farms, and offshore drilling platforms.

5. The method according to claim 1, wherein Before inputting the extracted area data into the deep learning image detection model to detect the type of the fixed target, the method further includes: Construct an image sample set of fixed targets at sea and label the fixed target type of each sample; Each sample is input into the deep learning image detection model for training, so that the model output continuously approaches the labeled fixed target type.

6. An electronic device, characterized in that: include: one or more processors; a memory for storing one or more programs, When the one or more programs are executed by the one or more processors, the one or more processors implement the large-area maritime fixed target rapid detection method based on multi-source satellite data as described in any one of claims 1-5.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method for rapid detection of large-area marine fixed targets based on multi-source satellite data as described in any one of claims 1 to 5 is implemented.

Images