Sea area island remote sensing information change monitoring method and system

Abstract

The invention relates to a sea area island remote sensing information change monitoring method and system, and relates to the technical field of remote sensing monitoring, and the method comprises the steps: obtaining a remote sensing image time sequence of a target sea area island, the remote sensing image time sequence comprises a plurality of remote sensing images, and each remote sensing image has an acquisition timestamp; collecting tide height information corresponding to the target sea area island at each collection timestamp, and water depth distribution information of the monitored sea area where the target sea area island is located; and performing correction processing on each remote sensing image according to the tidal height information and the water depth distribution information corresponding to each acquisition timestamp, and generating an island boundary contour change time sequence. The invention solves the problems that the traditional sea area island remote sensing monitoring adopts a single-moment image for analysis, the real coastline change and temporary boundary offset caused by tide cannot be distinguished, the boundary extraction precision is reduced due to shallow water area seabed reflection interference, the monitoring error is large, and the monitoring precision is low. The application precision requirements of sea area management, coastal zone protection and the like cannot be met.

Description

Technical Field

[0001] The present application relates to the field of remote sensing monitoring, and in particular to a method and system for monitoring changes in remote sensing information of sea areas and islands. Background Art

[0002] With the deepening of work such as sea area management and coastal protection, the accuracy of monitoring changes in remote sensing information of sea areas and islands has become a key technical problem.

[0003] Currently, traditional monitoring methods struggle to distinguish temporary shifts in island visual boundaries caused by tides from actual shoreline changes. Interference from seafloor reflections in shallow waters can significantly reduce boundary extraction accuracy. Furthermore, they often rely on single-moment image analysis, lacking a consistent benchmark for time-series monitoring. This not only reduces the value of island remote sensing monitoring data but also increases the complexity and cost of subsequent marine management and resource surveys, resulting in significant errors in monitoring results that fail to meet the accuracy requirements of practical applications. Summary of the Invention

[0004] In order to solve the above technical problems, the present application provides a method and system for monitoring changes in remote sensing information of sea areas and islands, which improves the current situation in traditional monitoring where the monitoring errors are large and it is difficult to meet the accuracy requirements of actual applications due to the inability to distinguish boundary offsets caused by tides, interference from shallow water reflections, and lack of a timing reference.

[0005] The embodiments of this application disclose the following technical solutions: In a first aspect, an embodiment of the present application provides a method for monitoring changes in remote sensing information of sea areas and islands, the method comprising: Acquire a time sequence of remote sensing images of the target sea area and islands, wherein the time sequence of remote sensing images includes a plurality of remote sensing images, each of the remote sensing images having an acquisition timestamp; Collecting the tidal height information of the target sea area and island at each collection time stamp, as well as the water depth distribution information of the monitoring sea area where the target sea area and island are located; Each of the remote sensing images is corrected and processed according to the tidal height information and the water depth distribution information corresponding to each of the acquisition timestamps to generate a time series of changes in the island boundary contour.

[0006] In a second aspect, an embodiment of the present application provides a system for monitoring changes in remote sensing information of sea areas and islands, the system comprising: An image time sequence acquisition module is used to acquire a remote sensing image time sequence of the target sea area and island, wherein the remote sensing image time sequence includes a plurality of remote sensing images, each of which has an acquisition timestamp; A time series tide level and water depth acquisition module is used to acquire the tidal height information of the target sea area and island at each acquisition time stamp, as well as the water depth distribution information of the monitoring sea area where the target sea area and island are located; The correction and time series generation module is used to correct each of the remote sensing images according to the tidal height information and the water depth distribution information corresponding to each of the acquisition timestamps, and generate a time series of changes in the island boundary contour.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages: This application proposes a method and system for monitoring changes in remote sensing information of sea areas and islands. By acquiring remote sensing image time series and environmental parameters, correcting remote sensing images by region, extracting boundary segments and splicing them to generate island boundary contours, and constructing a time series of changes in island boundary contours, accurate monitoring of changes in remote sensing information of sea areas and islands is achieved. First, the remote sensing image time series and the corresponding acquisition timestamps of the target sea area and islands are acquired, and the tidal height information of each timestamp and the water depth distribution information of the monitored sea area are synchronously acquired; then, the actual water depth distribution is calculated based on these environmental parameters, and the deep water and shallow water influence areas are divided accordingly. The first boundary segment is directly extracted in the deep water influence area, and the second boundary segment is extracted in the shallow water influence area after seabed reflection correction, and the two segments are spliced ​​together to form a single-time island boundary contour; finally, all contours are arranged in chronological order to generate a time series of changes in the island boundary contour, thereby achieving dynamic monitoring of island changes.

[0008] The technical solution of this application solves the problems of large errors and low accuracy caused by tidal offset, shallow water reflection interference, and lack of time series benchmark in traditional island monitoring by integrating the correction processing of remote sensing image time series and environmental parameters, extracting boundaries of different regions, constructing time series contours and quantitative analysis, and realizes high-precision dynamic monitoring of island boundary changes. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0010] Figure 1 A flow chart of a method for monitoring changes in remote sensing information of sea areas and islands provided in an embodiment of the present application; Figure 2 This is a structural diagram of a remote sensing information change monitoring system for sea areas and islands provided in an embodiment of the present application.

[0011] In the accompanying drawings, the components represented by the reference numerals are described as follows: Image time series acquisition module 01, time series tide level and water depth acquisition module 02, correction and time series generation module 03. DETAILED DESCRIPTION

[0012] The present application provides a method and system for monitoring changes in remote sensing information of sea areas and islands, which is used to solve the technical problems in the existing technology that it is difficult to distinguish between temporary offsets of the visual boundaries of islands and real coastline changes due to the influence of tides, the accuracy of boundary extraction is reduced due to interference from seabed reflections in shallow water areas, and the use of single-moment image analysis lacks a temporal monitoring consistency benchmark, which in turn causes large errors in island monitoring results and cannot meet the accuracy requirements of applications such as sea area management.

[0013] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without making creative efforts are within the scope of protection of this application.

[0014] In the description of this application, the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of the technical features indicated. Therefore, a feature specified as "first" or "second" may explicitly or implicitly include one or more of the described features. In the description of this application, "plurality" means two or more, unless otherwise specifically specified.

[0015] In the description of this application, the term "for example" is used to mean "used as an example, illustration or explanation". Any embodiment described as "for example" in this application is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is given to enable any person skilled in the art to implement and use the present invention. In the following description, details are listed for the purpose of explanation. It should be understood that a person of ordinary skill in the art can recognize that the present invention can be implemented without using these specific details. In other examples, well-known structures and processes are not elaborated in detail to avoid obscuring the description of the present invention with unnecessary details. Therefore, the present invention is not intended to be limited to the embodiments shown, but is consistent with the widest scope consistent with the principles and features disclosed in this application.

[0016] Example 1, as shown in the attached Figure 1 As shown, the present application provides a method for monitoring changes in remote sensing information of sea areas and islands, the method comprising the following steps: S110: Acquire a time sequence of remote sensing images of the target sea area and islands, wherein the time sequence of remote sensing images includes a plurality of remote sensing images, each of which has an acquisition timestamp; In an embodiment of the present application, in the scenario of monitoring changes in remote sensing information of sea areas and islands, in order to obtain coherent and accurate remote sensing image time series data, it is necessary to associate the image library with the island identification and filter it according to the time parameters to establish a basic data sequence for time series analysis.

[0017] Specifically, the unique island identifier of the target sea area island is first obtained through the preset geographic information database, and the corresponding remote sensing image library is retrieved based on this identifier to ensure an accurate match between the data source and the detection object.

[0018] At the same time, the acquisition time interval and monitoring time period set by the user are received as the time basis for image screening. Among them, the monitoring time period includes the start time and the end time.

[0019] Furthermore, a plurality of remote sensing images that meet the conditions are screened out from the remote sensing image library according to the above time parameters, and the acquisition time of each remote sensing image is synchronously extracted as a timestamp to form a remote sensing image time series containing multi-time data.

[0020] This step provides consistent basic image data with time series for subsequent remote sensing image correction based on tide and water depth data through precise association of island identification with remote sensing image database and image screening based on time reference, thus ensuring the continuity and accuracy of remote sensing information change monitoring.

[0021] Step S110 of the method provided in the embodiment of the present application includes: Obtaining the island identification of the target sea area and retrieving the corresponding remote sensing image library based on the island identification; Receiving a collection time interval and a monitoring time period set by a user, wherein the monitoring time period includes a start time and an end time; According to the acquisition time interval, the start time and the end time, multiple remote sensing images are screened from the remote sensing image library, and the acquisition time of each remote sensing image is synchronously extracted as the acquisition timestamp of each remote sensing image.

[0022] In the embodiment of the present application, in order to obtain a remote sensing image time series that meets the monitoring requirements, it is necessary to accurately associate the remote sensing image library through island identification, and filter the data in combination with the time parameters set by the user to establish a basic image sequence for time series analysis to ensure the consistency and accuracy of subsequent correction processing.

[0023] Specifically, the island identifiers of the target sea area and islands are first obtained through a preset geographic information database. The identifiers are the island's unique number, geographic coordinate code, etc., which are the core index of the remote sensing image library.

[0024] For example, the island identifier of an island in a target sea area is "HD-2023-015". Through this identifier, all remote sensing images related to the island in the remote sensing image library can be directly retrieved to ensure that the data source strictly matches the monitored object and avoid errors caused by confusion of island information.

[0025] Among them, the remote sensing image library is pre-built based on the association mapping relationship between island identification and remote sensing images. The library stores remote sensing images of the island and surrounding waters collected by different satellites and different sensors at different times, and is classified and indexed according to attributes such as acquisition time and image resolution.

[0026] Furthermore, the user-set collection time interval and monitoring time period are received, wherein the collection time interval can be set according to the monitoring accuracy requirement, such as every 3 days, every 1 week or once a month.

[0027] Secondly, the monitoring period must clearly define the start and end time, for example, "January 1, 2024 to December 31, 2024." These time parameters are key to screening remote sensing images and directly determine the temporal density and coverage of the remote sensing image time series.

[0028] For example, if the user sets the collection time interval to 10 days and the monitoring period is from March 1, 2024 to June 1, 2024, it is necessary to filter out remote sensing images within the monitoring period and with a collection time interval of about 10 days from the remote sensing image library, such as remote sensing images at time points such as March 1, March 11, March 21, and April 1, to ensure the regularity of the time series.

[0029] Furthermore, based on the set acquisition time interval, monitoring start time and end time, multiple remote sensing images that meet the time parameter conditions are screened out from the retrieved remote sensing image library.

[0030] Specifically, in the image screening process, the time retrieval interface of the remote sensing image library is first called, the monitoring start time and end time are input, and all remote sensing images within the monitoring time period are preliminarily screened out.

[0031] Furthermore, the preliminary screening results are filtered twice according to the set collection time interval. For example, if the collection time interval is 7 days, images are selected every 7 days starting from the starting time to ensure that the images are evenly distributed in time and meet the monitoring frequency requirements.

[0032] At the same time, the integrity and availability of remote sensing images will be verified synchronously during the screening process, and images with severe noise, cloud cover exceeding the threshold or data damage will be eliminated. Finally, multiple remote sensing images that meet the conditions will be retained, and the acquisition time of each image will be extracted as the timestamp. The images are arranged in chronological order to form a continuous remote sensing image time series.

[0033] For example, if the user sets the monitoring period from May 1, 2024, to August 1, 2024, and the collection interval is 15 days, remote sensing images within the monitoring period are first retrieved from the remote sensing image library. After verification, the image of May 16, which was completely covered by clouds due to heavy rain, is removed. Finally, a total of 6 valid images are selected, namely May 1, May 31, June 15, June 30, July 15, and August 1. Their timestamps correspond to the collection time of each image, and they are arranged in order to form a complete remote sensing image time series.

[0034] Finally, the remote sensing image time series obtained through the above steps contains multiple valid remote sensing images with precise timestamps, which lays a solid data foundation for subsequent correction processing based on tidal height and water depth distribution information and generation of island boundary contour change time series, ensuring the time series consistency and data reliability of the entire monitoring process.

[0035] S120: Collecting tidal height information corresponding to each collection time stamp of the target sea area and the water depth distribution information of the monitoring sea area where the target sea area and the island are located; In the embodiment of the present application, in the scenario of monitoring changes in remote sensing information of sea areas and islands, in order to accurately obtain tide and water depth data for image correction, it is necessary to combine the characteristics of the islands and the attributes of the remote sensing images to determine the monitoring range, and then collect key information in a targeted manner to provide data support for subsequent remote sensing correction processing.

[0036] Specifically, firstly, the island feature information of the target sea area islands and the remote sensing image resolution of each remote sensing image are obtained.

[0037] Among them, the island feature information includes the island coordinate position and island area. These island feature information are the basis for locating the monitoring area and calculating the buffer range; the remote sensing image resolution directly affects the setting of the buffer distance to ensure that the collected water depth data is compatible with the remote sensing image accuracy.

[0038] The buffer zone is then determined based on the island area and remote sensing image resolution. This step uses quantitative calculations to define the boundaries of the monitored sea area, ensuring coverage of areas around the islands that may be affected by tides and water depths while avoiding data redundancy caused by an overly large area.

[0039] On this basis, the tidal height information corresponding to each acquisition time stamp is retrieved from the tidal database based on the coordinate position of the island to ensure that the water level status at each time point can be accurately matched, providing a data basis for eliminating the boundary offset caused by the tide.

[0040] At the same time, the monitoring sea area is determined based on the coordinate position of the island and the range of the buffer zone, and the water depth distribution information of the monitoring area is obtained from the ocean bathymetry database, laying the foundation for distinguishing between deep and shallow water areas and dealing with seabed reflection interference.

[0041] This step systematically collects tidal height information and water depth distribution information to accurately obtain key environmental parameters, providing reliable data support for the subsequent elimination of tidal and shallow water reflection interference and improving boundary extraction accuracy, thereby ensuring the accuracy of the monitoring results.

[0042] Step S120 in the method provided in the embodiment of the present application includes: Acquiring island feature information of the target sea area islands and remote sensing image resolution of each remote sensing image, wherein the island feature information includes the island coordinate position and the island area; Determining a buffer zone range based on the area of ​​the island and the resolution of the remote sensing image; Based on the coordinate position of the island, obtaining the tidal height information corresponding to each of the acquisition timestamps; The monitoring sea area is determined based on the coordinate position of the island and the range of the buffer zone, and the water depth distribution information of the monitoring sea area is obtained.

[0043] In the embodiment of the present application, in order to accurately collect tidal height and water depth distribution information for remote sensing image correction, it is necessary to first clarify the spatial attributes of the island and the image accuracy characteristics, provide a basis for demarcating a reasonable monitoring sea area, and ensure the pertinence and effectiveness of subsequent data collection.

[0044] Specifically, firstly, the island feature information of the target sea area islands and the remote sensing image resolution of each remote sensing image are obtained.

[0045] Among them, the coordinate position in the island feature information is the spatial benchmark for positioning tidal data and monitoring sea areas. For example, the coordinates of an island are 120°30′ east longitude and 30°15′ north latitude, which can accurately lock its position in geographic space.

[0046] Furthermore, island area reflects the size of the island and is a fundamental parameter for calculating the buffer zone. Remote sensing image resolution determines the image's ability to depict surface detail, directly influencing the setting of buffer distances to avoid data redundancy due to an overly large buffer zone or information loss due to a too small buffer zone.

[0047] Furthermore, the buffer zone range is determined based on the island area and remote sensing image resolution to accurately define the boundaries of the monitored sea area where water depth distribution information needs to be collected.

[0048] In the method provided in the embodiment of the present application, the step of “determining the buffer zone range based on the island area and the remote sensing image resolution” includes: Calculating an island equivalent radius based on the island area, and obtaining a first buffer distance from a buffer distance mapping table based on the island equivalent radius; Determining a second buffer distance based on the remote sensing image resolution, where the second buffer distance is a preset multiple of the remote sensing image resolution; Compare the first buffer distance with the second buffer distance, and take the larger value as the basic buffer radius; The buffer zone range is obtained by adding a safety margin to the basic buffer radius.

[0049] In the embodiment of the present application, in order to scientifically define the buffer zone range that is both suitable for the island scale and consistent with the accuracy of remote sensing images, it is necessary to perform multi-dimensional calculations based on the island area and the remote sensing image resolution to achieve accurate delineation of the monitored sea area and provide a reasonable spatial boundary for the effective collection of subsequent water depth distribution information.

[0050] Specifically, the equivalent radius of the island is first calculated based on the island area, that is, the irregular island shape is converted into a standardized circular scale parameter to facilitate the unified calculation of the buffer distance.

[0051] For example, if the area of ​​an island is 1256 square meters (π is 3.14), according to the circle area formula , the corresponding island equivalent radius can be derived rice.

[0052] Furthermore, the island equivalent radius is used as an index to match the first buffer distance in a preset buffer distance mapping table.

[0053] Among them, the buffer distance mapping table pre-stores standard buffer distances corresponding to different equivalent radii. The value setting is based on the typical range statistics affected by tides and water depths around the island. For example, an equivalent radius of 20 meters corresponds to a first buffer distance of 100 meters to ensure that the buffer distance matches the actual scale of the island.

[0054] Furthermore, a second buffer distance is determined based on the remote sensing image resolution. The second buffer distance is set to a preset multiple (e.g., 300 times) of the remote sensing image resolution. This design aims to ensure that the buffer zone covers the area corresponding to the smallest spatial detail recognizable in the image.

[0055] For example, if the remote sensing image resolution is 5 meters, the second buffer distance is 5×300=1500 meters, which effectively avoids the accuracy limitation of the remote sensing image causing the buffer range to be too small, thereby missing key landform information in shallow water areas.

[0056] Furthermore, the first buffer distance is compared with the second buffer distance, and the larger value is taken as the basic buffer radius to achieve a balance between the island scale and the accuracy of remote sensing images.

[0057] Specifically, when the island area is small but the remote sensing image resolution is high, the second buffer distance corresponding to the remote sensing image resolution is used as the primary buffer distance to ensure sufficient coverage of the surrounding sea area. When the island area is large but the remote sensing image resolution is low, the first buffer distance corresponding to the island scale is used as the primary buffer distance to avoid data redundancy caused by excessive coverage. For example, if the first buffer distance is 100 meters and the second buffer distance is 1500 meters, the basic buffer radius is 1500 meters.

[0058] Finally, a safety margin is added to the basic buffer radius to form the final buffer range. The safety margin is set to account for uncertainties such as tidal extremes and measurement errors, ensuring that areas that affect island boundary identification can still be fully covered in extreme cases.

[0059] For example, if 50 meters is taken as the safety margin, based on the basic buffer radius of 1500 meters and combined with the 50-meter safety margin, the final buffer range is 1550 meters (1500+50=1550).

[0060] Furthermore, after determining the buffer zone range, the tidal height information corresponding to each acquisition timestamp is accurately retrieved from the tidal database based on the island coordinate position.

[0061] Among them, since the tidal height has significant temporal and spatial specificity, the water level changes in the same sea area at different times will directly affect the visual boundary presentation of the island. Therefore, it is necessary to lock the coverage area of ​​the tidal monitoring station to which it belongs through the island coordinates, and then extract the tidal height information at the corresponding time based on the acquisition timestamp of each remote sensing image.

[0062] For example, the records of the tide monitoring station corresponding to the coordinates of a certain island show that the tide height was +0.8 meters at the time of the remote sensing image timestamp "2024-07-10-08:30". This data will serve as the key basis for subsequent correction of the boundary offset of the remote sensing image caused by the tide, to ensure that the remote sensing images with different timestamps remain consistent in the water level benchmark.

[0063] At the same time, combining the island coordinates and the defined buffer zone, the spatial boundaries of the monitoring sea area can be precisely delineated. That is, the circular area formed by the island coordinates as the center and the buffer zone as the radius is the monitoring sea area where water depth distribution information needs to be collected.

[0064] Furthermore, the water depth distribution information of the monitored sea area is retrieved from the ocean bathymetry database. This information usually includes the water depth values ​​of different coordinate points in the sea area, which can clearly reflect the changes in water depth gradient from the periphery of the island to the buffer boundary.

[0065] For example, the water depth corresponding to the coordinates of a certain point in the monitored sea area is 3 meters, and that of another point is 10 meters. This can be used to distinguish shallow water areas (such as water depth ≤ 5 meters) and deep water areas (such as water depth > 5 meters), providing basic data for the subsequent use of differentiated remote sensing image correction strategies for different water depth areas, effectively improving the accuracy of island boundary extraction.

[0066] This step achieves accurate definition of areas of different water depths by systematically acquiring water depth distribution information in the monitored sea area, laying a data foundation for the subsequent use of different correction strategies for shallow and deep water areas, helping to eliminate interference factors such as seabed reflections and improve the accuracy of island boundary extraction.

[0067] S130: Correcting each of the remote sensing images according to the tidal height information and the water depth distribution information corresponding to each of the acquisition timestamps to generate a time series of changes in the island boundary contour.

[0068] In an embodiment of the present application, in the scenario of monitoring changes in remote sensing information of sea areas and islands, in order to eliminate the interference of tides and shallow water reflections on remote sensing images, it is necessary to perform precise corrections based on environmental parameters associated with timestamps, and construct a time-consistent island boundary contour through regional processing to provide a reliable benchmark for dynamic change monitoring.

[0069] Specifically, first, based on the water depth distribution information and the tidal height information corresponding to each acquisition time stamp, the actual water depth distribution at each moment is calculated.

[0070] Among them, the water depth distribution information reflects the basic topographic characteristics of the monitored sea area, while the tidal height information reflects the water level fluctuations at different time points. The combination of the two can obtain the actual depth of seawater submergence at each time stamp.

[0071] Furthermore, the remote sensing images are corrected based on the actual water depth distribution corresponding to each acquisition time stamp. That is, by associating image pixels with actual water depth, environmental interference can be specifically eliminated.

[0072] Among them, for deep-water areas where the actual water depth exceeds the threshold affected by reflection, the water-land separation threshold is directly used to divide the boundary; for shallow-water areas where the actual water depth is within the reflection interference range, the water depth pixel corrector must be used to eliminate the influence of seabed reflection on the image pixel value before the island boundary contour is extracted.

[0073] On this basis, each remote sensing image in the remote sensing image time series is traversed to extract the actual water depth distribution of the corresponding timestamp, and the deep water influence area and shallow water influence area are determined.

[0074] At the same time, the first boundary segment is extracted in the deep water area, and the second boundary segment after reflection correction is extracted in the shallow water area. Through spatial splicing, a complete island boundary outline is formed at a single moment. All time series images are processed in this way to finally generate a time series of island boundary outline changes.

[0075] This step achieves the dual goals of tidal correction and water depth correction by deeply integrating environmental parameters with image processing, providing time-coherent data for accurately capturing the dynamic changes of the real boundaries of the island, and effectively supporting the demand for monitoring accuracy in scenarios such as sea area management.

[0076] Step S130 in the method provided in the embodiment of the present application includes: Calculating the actual water depth distribution corresponding to each acquisition time stamp based on the water depth distribution information and the tidal height information corresponding to each acquisition time stamp; Each of the remote sensing images is corrected according to the actual water depth distribution corresponding to each acquisition time stamp to obtain the island boundary contour corresponding to each acquisition time stamp, and generate the island boundary contour change time series.

[0077] In the embodiment of the present application, in order to accurately capture the real boundary changes of the island and solve problems such as boundary offset caused by tides, shallow water reflection interference and lack of timing reference, it is necessary to accurately correct the remote sensing image and then construct a time-consistent sequence of island boundary contour changes to eliminate the impact of environmental interference on boundary identification and provide a reliable basis for dynamic monitoring.

[0078] First, based on the water depth distribution information and the tidal height information corresponding to each acquisition time stamp, the actual water depth distribution corresponding to each acquisition time stamp is calculated.

[0079] Among them, the water depth distribution information reflects the basic topographic and water depth characteristics of the monitored sea area, for example, the baseline water depth of a certain area is 4 meters; and the tidal height information reflects the water level fluctuations at different time points. For example, the tidal height at a certain timestamp is +0.6 meters. The superposition of the two can obtain the actual water depth distribution (4+0.6=4.6 meters). This data is the core basis for distinguishing the boundaries between land and water and dividing the affected areas.

[0080] Furthermore, the remote sensing image is corrected according to the actual water depth distribution corresponding to each acquisition time stamp to obtain the island boundary contour at a single moment.

[0081] In the method provided in the embodiment of the present application, the step of “correcting each of the remote sensing images according to the actual water depth distribution corresponding to each acquisition time stamp to obtain the island boundary contour corresponding to each acquisition time stamp” includes: Traversing multiple remote sensing images in the remote sensing image time series to obtain a first remote sensing image; Extracting a first acquisition timestamp of the first remote sensing image, and obtaining a first actual water depth distribution according to the first acquisition timestamp; determining a deep water influence area and a shallow water influence area in the first remote sensing image according to the first actual water depth distribution; For the first remote sensing image, extract a first boundary segment of the target sea area and island in the deep water influence area, and extract a second boundary segment of the target sea area and island in the shallow water influence area; spatially stitching the first boundary segment and the second boundary segment to generate an island boundary outline corresponding to the first acquisition timestamp; The island boundary contours corresponding to each acquisition time stamp are obtained in a manner of obtaining the island boundary contour corresponding to the first acquisition time stamp.

[0082] In the embodiment of the present application, in order to accurately extract the island boundary contour at different time stamps, the remote sensing image needs to be processed in different regions based on the actual water depth distribution, and environmental interference is eliminated through differentiated boundary extraction methods to ensure the accuracy of the single-time contour and the temporal consistency of the multi-time contour, laying the foundation for generating the island boundary contour change time series.

[0083] Specifically, first, multiple remote sensing images in the remote sensing image time series are traversed, and any one of them is selected as the first remote sensing image.

[0084] Furthermore, a first acquisition timestamp of the first remote sensing image is extracted, and a corresponding first actual water depth distribution is retrieved according to the timestamp.

[0085] Among them, the first actual water depth distribution is key data that integrates the tidal height information at that moment and the basic water depth distribution information. For example, the first acquisition timestamp is "2024-08-01-14:00", and in the corresponding first actual water depth distribution, the water depth in some areas around the island is 6 meters, and in some areas it is 2 meters, providing a quantitative basis for subsequent regional division.

[0086] On this basis, the deep water influence area and the shallow water influence area are determined in the first remote sensing image according to the first actual water depth distribution.

[0087] Among them, the deep-water impact area refers to the sea area with a relatively deep actual water depth (such as >5 meters). This area is less affected by seabed reflections and has low difficulty in boundary identification; the shallow-water impact area refers to the sea area with a relatively shallow actual water depth (such as ≤5 meters). The seabed topography has a significant reflection effect on remote sensing signals, which can easily lead to blurred water-land boundaries.

[0088] For example, if the first actual water depth distribution shows that the water depth in the sea area east of the island is 7 meters, it is divided into a deep water influence area; if the water depth in the sea area west of the island is 3 meters, it is divided into a shallow water influence area.

[0089] Furthermore, boundary segments are extracted by region from the first remote sensing image to achieve accurate processing of regions with different environmental interference levels, thereby ensuring the pertinence and accuracy of boundary extraction.

[0090] In the method provided in the embodiment of the present application, the step of “extracting, for the first remote sensing image, a first boundary segment of the target sea area and island in the deep water influence area, and extracting a second boundary segment of the target sea area and island in the shallow water influence area” includes: Get the water-land separation threshold; In the deep-water impact area, pixel classification processing is performed on the first remote sensing image based on the land-water separation threshold, and a boundary line between seawater and target sea area islands is extracted to obtain the first boundary segment; In the shallow water impact area, after performing seabed reflection correction on the first remote sensing image, pixel classification processing is performed in combination with the water-land separation threshold, and the boundary line between the sea water and the target sea area and island is extracted to obtain the second boundary segment.

[0091] In the embodiment of the present application, in order to accurately extract island boundary segments in different water depth environments, it is necessary to perform differentiated processing based on regional interference characteristics to ensure the accuracy and compatibility of the two boundary segments, laying the foundation for subsequent island boundary contour splicing.

[0092] Specifically, first, a water-land separation threshold is obtained, which is a pixel value critical value obtained based on statistics of a large number of sample images and is used to distinguish the pixel features of seawater and islands and land.

[0093] For example, the land-water separation threshold is set to 0.55 (pixel value range 0-1). When the pixel value of the remote sensing image is greater than 0.55, it is judged as island land, and when it is less than 0.55, it is judged as sea water, providing a unified classification standard for boundary extraction.

[0094] Furthermore, a first boundary segment is extracted in the deep water influence area.

[0095] Among them, since the actual water depth in the deep-water impact area is relatively deep (actual water depth > 5 meters), the interference of seabed reflection on remote sensing images is extremely small, and the pixel values ​​of islands and seawater are significantly different (island pixel values ​​are concentrated in the range of 0.6-0.9, and seawater pixel values ​​are concentrated in the range of 0.2-0.4).

[0096] Therefore, pixel classification processing is performed on the first remote sensing image directly based on the land-water separation threshold to quickly and accurately divide the island land and sea water pixels in the deep-water influence area, providing a basis for extracting a clear first boundary segment.

[0097] Specifically, by traversing all remote sensing image pixels in the deep-water impact area, pixel values ​​greater than 0.55 are marked as land pixels, and those less than 0.55 are marked as seawater pixels. The boundary line between the two types of pixels is then extracted using the edge detection algorithm in the existing technology to obtain the first boundary fragment.

[0098] For example, after pixel classification in a certain deep-water impact area, the pixel distribution of island land and sea water shows a clear continuous boundary. The first boundary segment extracted by the edge detection algorithm is complete and burr-free, which can accurately reflect the real island boundary of the deep-water impact area.

[0099] At the same time, a second boundary segment is extracted in the shallow water impact area. Because the actual water depth in shallow water areas is relatively shallow (actual water depth ≤ 5 meters), the seabed topography significantly reflects the remote sensing signal, causing the pixel values ​​of islands and land to be lowered. Directly applying the land-water separation threshold will result in misclassification. Therefore, it is necessary to first perform seabed reflection correction processing on this shallow water impact area.

[0100] In the method provided in the embodiment of the present application, the step of "in the shallow water influence area, performing seabed reflection correction on the first remote sensing image, performing pixel classification processing in combination with the land-water separation threshold, and extracting the boundary between the seawater and the target sea area island to obtain the second boundary segment" includes: In the first remote sensing image, obtaining the actual water depth value and the original pixel value of each pixel point in the shallow water influence area; Performing seabed reflection correction on each pixel point in the shallow water influence area by a water depth pixel corrector to obtain a corrected pixel value, and replacing the original pixel value with the corrected pixel value; The first remote sensing image is subjected to pixel classification processing based on the corrected pixel value and the land-water separation threshold, and the boundary line between the seawater and the target sea area island is extracted to obtain the second boundary segment.

[0101] In an embodiment of the present application, in order to eliminate the interference of seabed reflections in shallow water areas on remote sensing images and ensure the accuracy of island boundary extraction in the area, it is necessary to correct the pixel values ​​through the "data acquisition-reflection correction-classification extraction" process and then divide the boundaries, so that the water-land boundary characteristics of the shallow water impact area and the deep water impact area maintain the same judgment criteria, providing compatible basic data for boundary fragment splicing.

[0102] Specifically, first, in the first remote sensing image, the actual water depth value and original pixel value of each pixel point in the shallow water influence area are obtained to establish a basic data pair for reflection correction, providing input parameters for the subsequent elimination of seabed reflection interference through the water depth pixel corrector, and ensuring that the corrected pixel value can truly reflect the surface properties.

[0103] Among them, the actual water depth value comes from the actual water depth distribution of the shallow water impact area, such as 2.1 meters and 3.7 meters, reflecting the actual depth of seawater coverage.

[0104] Furthermore, the original pixel value is the initial grayscale value of the pixel in the remote sensing image, such as 0.48 or 0.53. However, due to the influence of seabed reflection, these initial grayscale values ​​cannot truly distinguish between islands and seawater. For example, in a shallow water-affected area, the original pixel value of an island land pixel is only 0.52 due to interference from seabed reflection, which is very different from the seawater pixel value of 0.49, making it difficult to directly distinguish.

[0105] Therefore, in order to solve the problem that the original pixel values ​​in the shallow water influence area are interfered with by the seabed reflection and cannot truly distinguish between islands, land and sea water, it is necessary to use a water depth pixel corrector to perform seabed reflection correction on each pixel point in the shallow water influence area to eliminate the pixel value distortion caused by reflection interference, laying the foundation for subsequent accurate classification based on the water-land separation threshold.

[0106] In the method provided in the embodiment of the present application, the steps of constructing the “water depth pixel corrector” include: Collecting a sample actual water depth record set and a sample original pixel set, and annotating the sample actual water depth record set and the sample original pixel set to obtain a sample corrected water depth record set; The water depth pixel corrector is trained and generated by taking the sample actual water depth record set and the sample original pixel set as input and the sample corrected water depth record set as a supervisory signal.

[0107] In the embodiment of the present application, in order to construct a correction tool that can accurately eliminate seabed reflection interference in shallow water areas, a systematic sample collection, labeling and model training process is required to enable the water depth pixel corrector to have the ability to deduce the true surface pixel characteristics based on the actual water depth and original pixel values.

[0108] Specifically, the actual water depth record set of the samples and the original pixel set of the samples are first collected.

[0109] Among them, the actual water depth record set of the samples is the measured water depth data collected by high-precision ocean depth sounding equipment (such as multi-beam echo sounder) in different shallow water impact areas, which contains specific water depth values ​​corresponding to a large number of pixel points, such as 1.2 meters, 2.8 meters, 4.5 meters, etc. These data need to cover shallow water environments with different tidal states and seabed landform types to ensure the representativeness of the samples.

[0110] In addition, the sample original pixel set corresponds to the initial grayscale values ​​of these water depth sampling points in the remote sensing image, such as 0.42, 0.51, 0.55, etc., which directly reflects the original image characteristics affected by the seabed reflection interference.

[0111] For example, 10,000 sample points are collected in a sandy shoal area, of which 5,000 are land pixels around the island (the actual water depth is 0, covered by extremely shallow sea water), and 5,000 are pure seawater pixels. Their actual water depths and original pixel values ​​are recorded respectively to form a basic sample data set.

[0112] Furthermore, annotation is performed based on the sample actual water depth record set and the sample original pixel set to obtain a sample corrected pixel set.

[0113] In the labeling process, professionals combine field survey data to correct the original pixel values ​​and obtain the ideal pixel values ​​after eliminating reflection interference.

[0114] For example, a sample point is actually an island with an actual water depth of 0.3 meters. Because it is submerged at high tide, it is affected by the reflection of the seabed, resulting in the original pixel value being as low as 0.45. Referring to the similar land pixel characteristics without reflection interference, its corrected pixel value is marked as 0.72.

[0115] In addition, a certain sample point is actually pure seawater with an actual water depth of 3.2 meters. The original pixel value is affected by the seabed reflection, resulting in a high original pixel value of 0.53. Referring to the seawater pixel characteristics in the deep water impact area, its corrected pixel value is marked as 0.41.

[0116] Finally, after the labeling is completed, each sample point forms a complete data chain of "actual water depth value-original pixel value-corrected pixel value", providing a supervision basis for model training.

[0117] On this basis, the water depth pixel corrector is trained and generated with the sample actual water depth record set and the sample original pixel set as input and the sample corrected pixel set as the supervision signal.

[0118] Specifically, a deep neural network is used as the basic framework for model training. The model structure includes an input layer that receives the actual water depth value and the original pixel value, a hidden layer that extracts feature associations through multiple layers of neurons, such as the nonlinear relationship between water depth and reflection intensity, and an output layer that outputs the final corrected pixel value.

[0119] During the training process, the model parameters are continuously optimized through the back-propagation algorithm so that the mean square error between the output corrected pixel value and the sample corrected pixel set is controlled within the preset threshold (e.g. ≤0.02).

[0120] For example, 100,000 sets of sample data were used for training. After 50 rounds of iteration, the model's correction accuracy for the validation set reached more than 96%. That is, after inputting the actual water depth and original pixel value, the output corrected pixel value deviated very little from the ideal value manually labeled, which can meet the accuracy requirements of boundary extraction. At this time, the construction of the water depth pixel corrector is completed.

[0121] Finally, by constructing a training framework of "measured data-original features-ideal features", the water depth pixel corrector can accurately capture the interference patterns of seabed reflections in shallow water areas, providing an automated, high-precision correction tool for subsequent batch processing of shallow water area pixels in remote sensing images.

[0122] For example, a shallow water sample set contains a set of data with an actual water depth of 1.5 meters and an original pixel value of 0.48. However, due to the reflection of the sandy seabed, the land pixel value is pulled down to close to the seawater value. After manual annotation, the corresponding sample corrected pixel value is 0.73 to match the land pixel characteristics without reflection interference.

[0123] Furthermore, the sample is input into the trained water depth pixel corrector, and the corrector outputs a corrected pixel value of 0.72, which is consistent with the actual annotation value, and the accuracy meets the requirements.

[0124] In addition, when processing land pixels in shallow water areas with an actual water depth of 1.4 meters and an original pixel value of 0.49 in another remote sensing image, the water depth pixel corrector can automatically output a corrected pixel value of 0.71, which can effectively eliminate reflection interference and make the pixel significantly different from the surrounding seawater pixels, providing a reliable basis for subsequent classification processing based on the water-land separation threshold.

[0125] On this basis, the trained water depth pixel corrector is applied to the shallow water influence area of ​​the first remote sensing image, that is, all pixel points in the shallow water influence area are traversed, and the actual water depth value and original pixel value of each pixel are obtained one by one, and the corresponding corrected pixel value is obtained after inputting the water depth pixel corrector.

[0126] For example, the actual water depth of a certain pixel point is 2.1 meters and the original pixel value is 0.48. After processing by the water depth pixel corrector, the output corrected pixel value is 0.65; the actual water depth of another pixel point is 3.7 meters and the original pixel value is 0.53, and the output corrected pixel value is 0.51.

[0127] Furthermore, these corrected pixel values ​​were used to replace the original pixel values, so that the pixel characteristics of the shallow water area could be restored. For example, after correction, the land pixels were concentrated at 0.6-0.8, and the seawater pixels were concentrated at 0.4-0.5, and the difference between the two was significant.

[0128] Furthermore, based on the corrected pixel values, the first remote sensing image is pixel-classified in combination with a preset land-water separation threshold (eg, 0.55).

[0129] Specifically, pixel values ​​greater than 0.55 are labeled as island land pixels, while pixel values ​​less than 0.55 are labeled as seawater pixels. After classification, the boundary between the two types of pixels is extracted using the same edge detection algorithm used in the prior art. This boundary is the second boundary segment after eliminating reflection interference.

[0130] For example, after correction and classification of a shallow water influence area, the extracted second boundary segment is spatially continuous and smooth. Compared with the coastline measured in the field, the deviation is controlled within 1 pixel, and it naturally transitions with the first boundary segment of the deep water influence area at the junction, providing high-precision basic data for the subsequent splicing of the complete island boundary outline.

[0131] Furthermore, the first and second boundary segments are spatially spliced ​​to generate the island boundary outline corresponding to the first acquisition timestamp. This is achieved by using existing coordinate alignment and smooth transition algorithms to eliminate slight deviations at the junction of the two boundary segments, thereby ensuring the continuity and integrity of the overall outline.

[0132] Specifically, first extract the end point coordinate set of the first boundary segment and the start point coordinate set of the second boundary segment, and find the closest connection point pair through spatial coordinate matching. For example, the end point coordinates of the first segment ( , ) and the coordinates of the second segment's starting point ( , ), whose straight-line distance is less than 1 pixel unit.

[0133] Furthermore, the Bezier curve in the prior art is used to fit the transition path between the connection point pairs, so that the two boundaries are naturally merged to avoid sharp corners or breaks.

[0134] For example, there is a 0.3 pixel offset between the end point of the first boundary segment of the deep-water influence area of ​​an island and the starting point of the second boundary segment of the shallow-water influence area. After Bezier curve fitting, the spliced ​​boundary forms a continuous and smooth curve, and the overall deviation is controlled within 0.5 pixels, accurately reflecting the actual outline of the island.

[0135] On this basis, the island boundary outlines corresponding to each acquisition time stamp are obtained by following the same method as that corresponding to the first acquisition time stamp. That is, for each image in the remote sensing image sequence, the process of "extracting timestamps - obtaining actual water depth distribution - dividing deep and shallow water influence areas - extracting boundary segments by region - and spatially stitching" is repeated.

[0136] For example, for the second remote sensing image, the second actual water depth distribution corresponding to its second acquisition timestamp is extracted, the corresponding deep water and shallow water influence areas are divided, the third and fourth boundary fragments are extracted and spliced ​​respectively, and the island boundary contour corresponding to the second acquisition timestamp is obtained.

[0137] This process continues in this way until all time-series imagery is processed. The resulting multi-time island boundary contours, based on a unified calibration standard and splicing logic, are time-series consistent, providing accurate time-series data support for monitoring changes in remote sensing information about sea areas and islands.

[0138] On this basis, the island boundary contours corresponding to all acquisition timestamps were arranged in chronological order to form a coherent time series of island boundary contour changes. This time series, with time as the axis, fully records the morphological characteristics of the island boundary at different moments. By comparing the contour differences between adjacent timestamps, the dynamic change trend of the island boundary can be intuitively presented.

[0139] For example, in the time series contours of a certain island for three consecutive months, the boundary of the shallow water area on the east side advanced 5 meters toward the sea, while the boundary of the deep water area on the west side remained basically stable. This change can be directly quantified through the time series of changes in the island boundary contour, which will help subsequent marine area management departments to conduct dynamic assessments of island resources and coastal protection planning to ensure the balance between island ecosystems and resource development.

[0140] The embodiments of the present application achieve the following technical effects through the above specific implementation methods: This application proposes a method for monitoring changes in remote sensing information of sea areas and islands. First, a time series of remote sensing images of target sea areas and islands is obtained. Based on the island identifier, the corresponding remote sensing image library is retrieved and the remote sensing images with acquisition timestamps are filtered according to the time parameters set by the user. Then, the tidal height information and the water depth distribution information of the monitored sea area corresponding to each time stamp are collected. The buffer range is determined by the island characteristics and the image resolution, and the monitoring area is accurately delineated. Then, the actual water depth distribution is calculated based on the water depth distribution information and the tidal height information. Based on this, the image is corrected for each region. The first boundary segment is directly extracted from the deep-water affected area using the water-land separation threshold. The second boundary segment is extracted from the shallow-water affected area after the reflection interference is eliminated by the water depth pixel corrector, and the two segments are spliced ​​to form the single-time island boundary contour. Finally, all contours are arranged in chronological order to generate a time series of island boundary contour changes, so as to achieve accurate capture and time-series presentation of the dynamic changes of the island boundary.

[0141] The method provided in the embodiment of the present application adopts the technical solution of "image time series acquisition - environmental parameter collection - regional correction and boundary extraction - time series contour construction", which integrates technical means such as precise correction of tide and water depth data, differentiated boundary extraction of deep and shallow water areas, and time series analysis based on unified standards. It solves the problems of large errors and low accuracy caused by tidal offset, shallow water reflection interference, and lack of time series benchmark in traditional monitoring, and realizes high-precision dynamic monitoring of changes in remote sensing information of sea areas and islands.

[0142] Example 2, as shown in the attached Figure 2As shown, based on the inventive concept of a method for monitoring changes in remote sensing information of sea areas and islands provided in Example 1, this application also provides a system for monitoring changes in remote sensing information of sea areas and islands, specifically comprising: Image time sequence acquisition module 01 is used to obtain a remote sensing image time sequence of the target sea area and island, wherein the remote sensing image time sequence includes multiple remote sensing images, each of which has an acquisition timestamp; The time series tide level and water depth acquisition module 02 is used to acquire the tidal height information corresponding to each acquisition time stamp of the target sea area and the water depth distribution information of the monitoring sea area where the target sea area and island are located; The correction and time series generation module 03 is used to correct each of the remote sensing images according to the tidal height information and the water depth distribution information corresponding to each of the acquisition timestamps, and generate a time series of changes in the island boundary contour.

[0143] In one embodiment, the image timing acquisition module 01 is further configured to: Obtaining the island identification of the target sea area and retrieving the corresponding remote sensing image library based on the island identification; Receiving a collection time interval and a monitoring time period set by a user, wherein the monitoring time period includes a start time and an end time; According to the acquisition time interval, the start time and the end time, multiple remote sensing images are screened from the remote sensing image library, and the acquisition time of each remote sensing image is synchronously extracted as the acquisition timestamp of each remote sensing image.

[0144] In one embodiment, the time series tide level and water depth acquisition module 02 is further used to: Acquiring island feature information of the target sea area islands and remote sensing image resolution of each remote sensing image, wherein the island feature information includes the island coordinate position and the island area; Determining a buffer zone range based on the area of ​​the island and the resolution of the remote sensing image; Based on the coordinate position of the island, obtaining the tidal height information corresponding to each of the acquisition timestamps; The monitoring sea area is determined based on the coordinate position of the island and the range of the buffer zone, and the water depth distribution information of the monitoring sea area is obtained.

[0145] In one embodiment, the correction and timing generation module 03 is further configured to: Calculating the actual water depth distribution corresponding to each acquisition time stamp based on the water depth distribution information and the tidal height information corresponding to each acquisition time stamp; Each of the remote sensing images is corrected according to the actual water depth distribution corresponding to each acquisition time stamp to obtain the island boundary contour corresponding to each acquisition time stamp, and generate the island boundary contour change time series.

[0146] It should be noted that the order in which the embodiments of the present application are presented is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. Furthermore, the foregoing descriptions of specific embodiments of this specification are provided. Furthermore, the processes depicted in the accompanying drawings do not necessarily require the specific order or sequential sequence shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0147] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application should be included in the scope of protection of the present application.

[0148] This specification and drawings are merely illustrative of the present application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Obviously, those skilled in the art may make various modifications and variations to this application without departing from the scope of this application. Thus, this application is intended to include such modifications and variations as fall within the scope of this application and its equivalents.

Claims

1. A method for monitoring changes in remote sensing information of sea areas and islands, characterized in that: The method comprises: Acquire a time sequence of remote sensing images of the target sea area and islands, wherein the time sequence of remote sensing images includes a plurality of remote sensing images, each of the remote sensing images having an acquisition timestamp; Collecting the tidal height information of the target sea area and island at each collection time stamp, as well as the water depth distribution information of the monitoring sea area where the target sea area and island are located; Each of the remote sensing images is corrected and processed according to the tidal height information and the water depth distribution information corresponding to each of the acquisition timestamps to generate a time series of changes in the island boundary contour.

2. The method according to claim 1, characterized in that Acquire a time sequence of remote sensing images of the target sea area and islands, wherein the time sequence of remote sensing images includes multiple remote sensing images, each of which has an acquisition timestamp, including: Obtaining the island identification of the target sea area and retrieving the corresponding remote sensing image library based on the island identification; Receiving a collection time interval and a monitoring time period set by a user, wherein the monitoring time period includes a start time and an end time; According to the acquisition time interval, the start time and the end time, multiple remote sensing images are screened from the remote sensing image library, and the acquisition time of each remote sensing image is synchronously extracted as the acquisition timestamp of each remote sensing image.

3. The method according to claim 1, characterized in that Collecting the tidal height information of the target sea area island at each collection time stamp, and the water depth distribution information of the monitoring sea area where the target sea area island is located, including: Acquiring island feature information of the target sea area islands and remote sensing image resolution of each remote sensing image, wherein the island feature information includes the island coordinate position and the island area; Determining a buffer zone range based on the area of ​​the island and the resolution of the remote sensing image; Based on the coordinate position of the island, obtaining the tidal height information corresponding to each of the acquisition timestamps; The monitoring sea area is determined based on the coordinate position of the island and the range of the buffer zone, and the water depth distribution information of the monitoring sea area is obtained.

4. The method according to claim 3, characterized in that Determine the buffer zone based on the island area and the remote sensing image resolution, including: Calculating an island equivalent radius based on the island area, and obtaining a first buffer distance from a buffer distance mapping table based on the island equivalent radius; Determining a second buffer distance based on the remote sensing image resolution, where the second buffer distance is a preset multiple of the remote sensing image resolution; Compare the first buffer distance with the second buffer distance, and take the larger value as the basic buffer radius; The buffer zone range is obtained by adding a safety margin to the basic buffer radius.

5. The method according to claim 1, wherein Correcting each of the remote sensing images according to the tidal height information and the water depth distribution information corresponding to each of the acquisition timestamps to generate a time series of changes in the island boundary contour, including: Calculating the actual water depth distribution corresponding to each acquisition time stamp based on the water depth distribution information and the tidal height information corresponding to each acquisition time stamp; Each of the remote sensing images is corrected according to the actual water depth distribution corresponding to each acquisition time stamp to obtain the island boundary contour corresponding to each acquisition time stamp, and generate the island boundary contour change time series.

6. The method according to claim 1, wherein Correcting each of the remote sensing images according to the actual water depth distribution corresponding to each acquisition time stamp to obtain the island boundary contour corresponding to each acquisition time stamp includes: Traversing multiple remote sensing images in the remote sensing image time series to obtain a first remote sensing image; Extracting a first acquisition timestamp of the first remote sensing image, and obtaining a first actual water depth distribution according to the first acquisition timestamp; determining a deep water influence area and a shallow water influence area in the first remote sensing image according to the first actual water depth distribution; For the first remote sensing image, extract a first boundary segment of the target sea area and island in the deep water influence area, and extract a second boundary segment of the target sea area and island in the shallow water influence area; spatially stitching the first boundary segment and the second boundary segment to generate an island boundary outline corresponding to the first acquisition timestamp; The island boundary contours corresponding to each acquisition time stamp are obtained in a manner of obtaining the island boundary contour corresponding to the first acquisition time stamp.

7. The method according to claim 6, characterized in that For the first remote sensing image, extracting a first boundary segment of the target sea area and island in the deep water influence area, and extracting a second boundary segment of the target sea area and island in the shallow water influence area, including: Get the water-land separation threshold; In the deep-water impact area, pixel classification processing is performed on the first remote sensing image based on the land-water separation threshold, and a boundary line between seawater and target sea area islands is extracted to obtain the first boundary segment; In the shallow water impact area, after performing seabed reflection correction on the first remote sensing image, pixel classification processing is performed in combination with the water-land separation threshold, and the boundary line between the sea water and the target sea area and island is extracted to obtain the second boundary segment.

8. The method according to claim 7, characterized in that In the shallow water impact area, after performing seabed reflection correction on the first remote sensing image, pixel classification processing is performed in combination with the land-water separation threshold, and the boundary between the seawater and the target sea area island is extracted to obtain the second boundary segment, including: In the first remote sensing image, obtaining the actual water depth value and the original pixel value of each pixel point in the shallow water influence area; Performing seabed reflection correction on each pixel point in the shallow water influence area by a water depth pixel corrector to obtain a corrected pixel value, and replacing the original pixel value with the corrected pixel value; The first remote sensing image is subjected to pixel classification processing based on the corrected pixel value and the land-water separation threshold, and the boundary line between the seawater and the target sea area island is extracted to obtain the second boundary segment.

9. The method according to claim 8, characterized in that The steps of constructing the water depth pixel corrector include: Collecting a sample actual water depth record set and a sample original pixel set, and annotating the sample actual water depth record set and the sample original pixel set to obtain a sample corrected water depth record set; The water depth pixel corrector is trained and generated by taking the sample actual water depth record set and the sample original pixel set as input and the sample corrected water depth record set as a supervisory signal.

10. A remote sensing information change monitoring system for sea areas and islands, characterized in that: The system is used to execute the method for monitoring changes in remote sensing information of sea areas and islands according to any one of claims 1 to 9, and the system comprises: An image time sequence acquisition module is used to acquire a remote sensing image time sequence of the target sea area and island, wherein the remote sensing image time sequence includes a plurality of remote sensing images, each of which has an acquisition timestamp; A time series tide level and water depth acquisition module is used to acquire the tidal height information of the target sea area and island at each acquisition time stamp, as well as the water depth distribution information of the monitoring sea area where the target sea area and island are located; The correction and time series generation module is used to correct each of the remote sensing images according to the tidal height information and the water depth distribution information corresponding to each of the acquisition timestamps, and generate a time series of changes in the island boundary contour.

Images