Coastline intelligent extraction method and system based on environmental parameter self-adaption

Abstract

The invention provides a coastline intelligent extraction method and system based on environmental parameter self-adaption, and the method comprises the steps: obtaining remote sensing image data of a target region of a to-be-extracted coastline and tidal water level data of an imaging time phase of the target region, and carrying out the preprocessing of the remote sensing image data; carrying out inversion calculation on the suspended sediment concentration SSC of the target area based on the preprocessed remote sensing image data, and dynamically selecting a water body index to calculate a self-adaptive water body index graph corresponding to the target area; performing threshold segmentation on the self-adaptive water body index graph to obtain a preliminary water body mask graph; according to the tidal water level data, judging the tidal period of the imaging time phase, and dynamically selecting morphological operation based on the tidal period to carry out boundary optimization on the initial water body mask pattern to obtain an optimized water body mask pattern; a coastline is extracted from the optimized water body mask map and post-processing is carried out, and a final target area coastline is obtained; according to the method, high-precision and high-robustness coastline automatic extraction in a complex near-shore dynamic change environment can be realized.

Description

Technical Field

[0001] This invention relates to the field of remote sensing image processing technology, and more specifically, to a method and system for intelligent extraction of coastlines based on adaptive environmental parameters. Background Technology

[0002] As a key interface for land-sea interaction, the accurate extraction of coastlines is crucial for coastal resource management, ecological environment monitoring, and disaster risk assessment. Remote sensing technology, with its advantages of large-scale, periodic observation, has become the primary method for coastline extraction. Among these methods, threshold segmentation based on water body indices (such as NDWI, MNDWI, and IWI) is widely used due to its clear principles and computational efficiency. However, existing technologies face significant challenges in practical applications, especially in complex nearshore environments: 1) Environmental parameter sensitivity leads to the failure of water body indices: The suspended sediment concentration (SSC) of nearshore waters varies drastically in time and space. High sediment concentration will significantly change the spectral characteristics of water bodies, making it difficult for single or fixed combinations of water body indices to maintain high accuracy in different sediment concentration regions. This often results in water body misclassification (such as misclassifying high turbidity water bodies as land) or omissions.

[0003] 2) Tidal dynamics interfere with boundary morphology: The tidal flat area is affected by the periodic flooding and exposure of the tides, and its land-water boundary morphology changes significantly with the tide level. Traditional morphological post-processing methods (such as opening and closing operations) usually use fixed structuring elements and operation sequences, which are difficult to adapt to the boundary detail changes caused by the tides (such as blurred boundaries during high tide and broken boundaries during low tide), often resulting in problems such as broken, jagged, or lost details in the extracted coastline.

[0004] 3) Parameter selection relies on experience and lacks adaptability: Existing methods often rely on expert experience or repeated experiments for specific scenarios in the selection of water indexes, threshold settings, and morphological processing parameters (structural element type, size, and operation sequence). They lack a dynamic response mechanism to changes in environmental parameters (such as SSC and tide level). This static and fixed processing mode is difficult to adapt to the high heterogeneity and dynamism of the nearshore environment, which limits the improvement of the accuracy and automation level of coastline extraction. Summary of the Invention

[0005] To overcome the problems of low water body identification accuracy, poor boundary morphology optimization effect, and parameter selection dependence on experience and lack of adaptability caused by significant differences in suspended sediment concentration in nearshore waters and tidal dynamic changes in the prior art, this invention provides a coastline intelligent extraction method and system based on environmental parameter adaptation, which can realize high-precision and high-robust automated coastline extraction under complex nearshore dynamic change environment.

[0006] To solve the above technical problems, the technical solution of the present invention is as follows: An intelligent coastline extraction method based on environmental parameter adaptation, comprising the following steps: S1: Obtain the remote sensing image data of the target area of the coastline to be extracted and the tidal water level data at the imaging time phase, and preprocess the remote sensing image data; S2: Invert and calculate the suspended sediment concentration SSC of the target area based on the preprocessed remote sensing image data, and dynamically select a water body index according to the suspended sediment concentration SSC, and calculate the adaptive water body index map corresponding to the target area; S3: Perform threshold segmentation on the adaptive water body index map to obtain a preliminary water body mask map; S4: According to the tidal water level data, judge the tidal period at the imaging time phase, and dynamically select a morphological operation based on the tidal period to optimize the boundary of the preliminary water body mask map to obtain an optimized water body mask map; S5: Extract the coastline from the optimized water body mask map and perform post-processing to obtain the final coastline of the target area.

[0007] Preferably, in the step S1, the remote sensing image data is specifically multi-spectral remote sensing image data, and the spectral bands at least include a blue light band, a green light band, a red light band, a near-infrared band and a short-wave infrared band; The preprocessing at least includes any one or more of radiometric calibration, atmospheric correction, geometric precise correction and cloud masking.

[0008] Preferably, in the step S2, the suspended sediment concentration SSC of the target area is calculated based on a water color remote sensing inversion model, expressed as:

[0009] where is the remote sensing reflectance obtained from the preprocessed remote sensing image data; and are the red light wavelength and the green light wavelength corresponding to the preprocessed remote sensing image data respectively; 、 and are the first to third empirical parameters respectively. <00?00038>Preferably, in the step S2, dynamically selecting a water body index according to the suspended sediment concentration SSC includes: When SSC ≤ the first preset threshold, select the NDWI water body index; [[ID=?6]]When the first preset threshold < SSC ≤ the second preset threshold, select the MNDWI water body index; When SSC > second predetermined threshold, a merged water index is selected, wherein the merged water index is a linear weighted fusion of the IWI water index and the NDWI water index.

[0011] Preferably, in step S3, the Otsu threshold segmentation algorithm is used to perform binary segmentation of the adaptive water index map into water and non-water bodies to obtain a preliminary water mask map.

[0012] Preferably, in step S4, when the tidal level is greater than the third preset threshold, it is determined to be the high tide period, and boundary optimization is performed using closed-loop operation; When the tidal level is less than or equal to the third preset threshold, it is determined to be a low tide period, and the boundary optimization is performed using opening operation.

[0013] Preferably, the closing operation uses a rhombus structuring element for boundary optimization, and the opening operation uses a square structuring element for boundary optimization.

[0014] Preferably, in step S5, the optimized water body mask is converted into a vector polygon layer, and the water body boundary line in the vector polygon layer is extracted as a preliminary coastline using an edge detection algorithm; The preliminary coastline is then post-processed to obtain the final target area coastline.

[0015] Preferably, in step S5, the post-processing includes: fragment removal, smoothing based on B-spline curve fitting algorithm, topology anomaly detection and topology repair.

[0016] This invention also provides a coastline intelligent extraction system based on adaptive environmental parameters, which, using the above-described method, includes: Data acquisition module: used to acquire remote sensing image data of the target area of ​​the coastline to be extracted and the tidal water level data of the imaging time phase, and to preprocess the remote sensing image data. Water index calculation module: used to invert and calculate the suspended sediment concentration (SSC) of the target area based on the preprocessed remote sensing image data, and dynamically select the water index according to the suspended sediment concentration (SSC) to calculate the adaptive water index map corresponding to the target area. Preliminary segmentation module: used to perform threshold segmentation on the adaptive water index map to obtain a preliminary water mask map; Boundary optimization module: used to determine the tidal period of the imaging phase based on the tidal water level data, and to perform boundary optimization on the preliminary water body mask map based on the dynamic selection of morphological operations according to the tidal period, so as to obtain the optimized water body mask map; Coastline extraction module: used to extract the coastline from the optimized water mask map and perform post-processing to obtain the final target area coastline.

[0017] Compared with the prior art, the beneficial effects of the technical solution of the present invention are: This invention provides a method and system for intelligent coastline extraction based on adaptive environmental parameters. First, remote sensing image data of the target area and tidal level data at the imaging time are acquired, and the remote sensing image data is preprocessed. Second, the suspended sediment concentration (SSC) of the target area is calculated based on the preprocessed remote sensing image data, and a water index is dynamically selected based on the SSC to calculate an adaptive water index map corresponding to the target area. Next, the adaptive water index map is thresholded to obtain a preliminary water mask map. Then, based on the tidal level data, the tidal period at the imaging time is determined, and morphological operations are dynamically selected based on the tidal period to optimize the boundaries of the preliminary water mask map, resulting in an optimized water mask map. Finally, the coastline is extracted from the optimized water mask map and post-processed to obtain the final target area coastline.

[0018] The beneficial effects of this invention are as follows: 1) Achieving intelligent dynamic selection of water body indices: This invention can adaptively select the optimal water body index based on the real-time predicted suspended sediment concentration, effectively solving the problem of the failure of traditional single indices under high turbidity water bodies, and significantly improving the identification accuracy in complex water environments.

[0019] 2) Establishing a boundary optimization mechanism driven by tidal water level: This invention dynamically adjusts the type, size and operation sequence of morphological processing (opening operation and closing operation) based on the tidal water level height (such as high tide and low tide), which can effectively suppress noise, enhance details, and significantly improve the continuity, smoothness and detail preservation ability of the tidal flat boundary under the action of tides.

[0020] 3) Enhancing the automation and robustness of coastline extraction: This invention internalizes the dynamic changes of key environmental parameters such as suspended sediment concentration and tidal phase into decision-making basis, reduces human experience intervention, and realizes the intelligent and adaptive process of coastline extraction in complex nearshore environments, thereby improving the adaptability and robustness of the whole solution under different regional and temporal conditions. Attached Figure Description

[0021] Figure 1 This is a flowchart of a method for intelligent coastline extraction based on adaptive environmental parameters, as provided in Example 1.

[0022] Figure 2 This is an overview map of the study area provided in Example 2.

[0023] Figure 3 This is a spatial distribution map of suspended sediment concentration provided in Example 2.

[0024] Figure 4This is a comparison diagram of the water body region segmentation effect before and after dynamic decision-making based on water body index provided in Example 2.

[0025] Figure 5 This is a comparison image of the morphological processing details provided in Example 2 (climax).

[0026] Figure 6 This is a comparison image of the morphological processing details provided in Example 2 (low tide period).

[0027] Figure 7 This is a comparison chart of the accuracy of the method provided in Example 2 and the coastline extraction based on high-resolution imagery.

[0028] Figure 8 This is a diagram of an intelligent coastline extraction system architecture based on adaptive environmental parameters, as provided in Example 3. Detailed Implementation

[0029] The accompanying drawings are for illustrative purposes only and should not be construed as limiting the scope of this application. To better illustrate this embodiment, some parts in the accompanying drawings may be omitted, enlarged, or reduced, and do not represent the actual product dimensions; It will be understood by those skilled in the art that certain well-known structures and their descriptions may be omitted in the accompanying drawings.

[0030] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0031] Example 1 like Figure 1 As shown, this embodiment provides a method for intelligent coastline extraction based on adaptive environmental parameters, including the following steps: S1: Acquire remote sensing image data of the target area of ​​the coastline to be extracted and its tidal water level data at the imaging time, and preprocess the remote sensing image data. S2: Based on the preprocessed remote sensing image data, the suspended sediment concentration (SSC) of the target area is calculated, and the water index is dynamically selected according to the suspended sediment concentration (SSC) to calculate the adaptive water index map corresponding to the target area. S3: Perform threshold segmentation on the adaptive water index map to obtain a preliminary water mask map; S4: Based on the tidal water level data, determine the tidal period of the imaging phase, and perform boundary optimization on the preliminary water body mask map by dynamically selecting morphological operations based on the tidal period to obtain the optimized water body mask map. S5: Extract the coastline from the optimized water mask map and perform post-processing to obtain the final target area coastline.

[0032] In the specific implementation process, the remote sensing image data of the target area of ​​the coastline to be extracted and the tidal water level data of the imaging time phase are first obtained, and the remote sensing image data are preprocessed. Secondly, the suspended sediment concentration (SSC) of the target area is calculated based on the preprocessed remote sensing image data, and the water index is dynamically selected according to the suspended sediment concentration (SSC) to calculate the adaptive water index map corresponding to the target area. Next, threshold segmentation is performed on the adaptive water index map to obtain a preliminary water mask map; Then, based on the tidal water level data, the tidal period of the imaging phase is determined, and morphological operations are dynamically selected based on the tidal period to optimize the boundary of the preliminary water mask map, resulting in the optimized water mask map. Finally, the coastline is extracted from the optimized water mask image and post-processed to obtain the final target area coastline. This method enables high-precision and robust automated extraction of coastlines under complex nearshore dynamic environments.

[0033] Example 2 This embodiment provides a method for intelligent coastline extraction based on adaptive environmental parameters, including the following steps: S1: Acquire remote sensing image data of the target area of ​​the coastline to be extracted and its tidal water level data at the imaging time, and preprocess the remote sensing image data. S2: Based on the preprocessed remote sensing image data, the suspended sediment concentration (SSC) of the target area is calculated, and the water index is dynamically selected according to the suspended sediment concentration (SSC) to calculate the adaptive water index map corresponding to the target area. S3: Perform threshold segmentation on the adaptive water index map to obtain a preliminary water mask map; S4: Based on the tidal water level data, determine the tidal period of the imaging phase, and perform boundary optimization on the preliminary water body mask map by dynamically selecting morphological operations based on the tidal period to obtain the optimized water body mask map. S5: Extract the coastline from the optimized water mask map and perform post-processing to obtain the final target area coastline. In step S1, the remote sensing image data is specifically multispectral remote sensing image data, and the spectral bands include at least the blue light band, green light band, red light band, near-infrared band and short-wave infrared band. The preprocessing includes at least one or more of the following: radiometric calibration, atmospheric correction, geometric fine correction, and declouding masking. In step S2, the suspended sediment concentration SSC of the target area is calculated based on the water color remote sensing inversion model, and is expressed as follows:

[0034] Among them, is the remote sensing reflectance obtained from the preprocessed remote sensing image data; and are the red light wavelength and green light wavelength corresponding to the preprocessed remote sensing image data respectively; , and are the first to third empirical parameters respectively; In the step S2, dynamically selecting a water body index according to the suspended sediment concentration SSC includes: When SSC ≤ the first preset threshold, select the NDWI water body index; When the first preset threshold < SSC ≤ the second preset threshold, select the MNDWI water body index; When SSC > the second preset threshold, select a fused water body index, where the fused water body index is a linear weighted fusion of the IWI water body index and the NDWI water body index; In the step S3, adopt the Otsu threshold segmentation algorithm to perform binary segmentation of water body and non - water body on the adaptive water body index map to obtain a preliminary water body mask map; In the step S4, when the tidal level is greater than the third preset threshold, it is determined as the high - tide period, and closing operation is used for boundary optimization; When the tidal level is less than or equal to the third preset threshold, it is determined as the low - tide period, and opening operation is used for boundary optimization; The closing operation uses a diamond - shaped structuring element for boundary optimization, and the opening operation uses a square - shaped structuring element for boundary optimization; In the step S5, convert the optimized water body mask map into a vector polygon layer, and use an edge detection algorithm to extract the water body boundary line in the vector polygon layer as the preliminary coastline; Perform post - processing on the preliminary coastline to obtain the final target area coastline; In the step S5, the post - processing includes: speckle removal, smoothing processing based on the B - spline curve fitting algorithm, topological anomaly checking and topological repair.

[0035] In the specific implementation process, this embodiment selects the inner - bay coastal zone of a certain province as the research area, which covers multiple municipal districts, such as Figure 2 shown; at the same time, 10 tide gauges A1 - A10 located in different administrative regions are selected, covering all geomorphic types such as muddy tidal flats, sandy shorelines, and mangrove wetlands.

[0036] The specific implementation steps are as follows: (1) Multi - source data collaborative acquisition and pre - processing: Remote sensing data acquisition: Acquire multispectral remote sensing images of the target area. In this embodiment, the key bands used and their center wavelengths / ranges include: Blue light band (Band 2): 450-510 nm; Green light band (Band 3): 530-590 nm; Red band (Band 4): 640-670 nm (for IWI calculations); Near-infrared band (Band 5): 850-880 nm (for NDWI calculations); Shortwave infrared band (Band 6): 1570-1650 nm (used for MNDWI calculations); Tidal data acquisition: Measured tidal level data for the corresponding imaging phases in the study area were acquired synchronously through the National Marine Science Data Center. In this embodiment, 10 representative tide gauge stations covering different shoreline types within the study area were selected, and their specific information is shown in Table 1. Table 1. Tidal Data Information for the Study Area

[0037] Preprocessing: The acquired remote sensing images are subjected to necessary preprocessing, including radiometric calibration, atmospheric correction, geometric fine correction, and cloud and cloud shadow masking using the QA band; (2) Calculation of the adaptive water index for environmental parameters: Suspended sediment concentration (SSC) inversion: The distribution of suspended sediment concentration in the study area was calculated based on the water color remote sensing inversion model. The SSC distribution map is shown below. Figure 3 As shown, the formula for calculating SSC is as follows:

[0038] in, This is the atmospheric-corrected remote sensing reflectance. =660±10nm, =560±10nm, , and These are the first to third empirical parameters, and in this embodiment, they are taken as follows: =285.7, =1.32, =5.4.

[0039] SSC Classification and Dynamic Index Decision-Making: Based on the SSC distribution map obtained through inversion, and according to the sediment concentration classification threshold defined in this method, the optimal water index is dynamically selected for each pixel (or region) in the image. In specific implementation, a random forest model can be further combined for dynamic decision-making. When SSC ≤ 30 mg / L, NDWI (Normalized Difference Water Index) is enabled to enhance the identification of clean water bodies by using the green and near-infrared bands:

[0040] In the formula, are the reflectances of the green band and the near-infrared band of the remote sensing image respectively; When 30 mg / L < SSC ≤ 50 mg / L, MNDWI (Modified Normalized Difference Water Index) is enabled to introduce the short-wave infrared band to suppress sediment interference:

[0041] In the formula, and are the reflectances of the green band and the short-wave infrared band of the remote sensing image respectively; When SSC > 50 mg / L, the IWI_M fusion water index is enabled to combine the advantages of IWI (Index-based Water Index) and NDWI: IWI_M = 0.8 × IWI + 0.2 × NDWI The calculation formula of IWI is:

[0042] In the formula, and and are the reflectances of the green band, the blue band and the red band of the remote sensing image respectively; According to the above decision results, an adaptive water index map of the entire study area is calculated pixel by pixel.

[0043] (3) Preliminary segmentation of water areas: For the adaptive water index map generated in step (2), the Otsu threshold segmentation algorithm is used to automatically determine the optimal segmentation threshold for binary segmentation of water bodies and non-water bodies, and a preliminary water body mask map (in raster format, the water body pixel value is 1 and the non-water body is 0) is obtained; as<00002​​​​​​​​​Based on the measured tide level data of each tide gauge station obtained in step (1), determine the tide period of the corresponding shoreline segment during imaging: If the measured tide level at a certain station is >2.0 m, the area affected by that station is determined to be in the high tide period. During the high tide period, the nearshore water body is flooded over a wide area, and small holes (such as isolated highlands in tidal channels) and narrow fractures (such as discontinuous boundaries caused by local siltation) are prone to appear on the tidal flats. Therefore, for the preliminary water body mask part corresponding to this area, a 5×5 rhombic structural element is used for closing operation (expansion followed by erosion) to suppress small holes and narrow fractures, smooth the boundaries and fill small gaps. Figure 5 This section showcases a comparison of morphological details during the climax phase (for ease of visual comparison). Figure 5 (The raster mask has been converted to a vector polygon): The left side shows the effect before processing, with multiple small holes with a diameter of less than 2 pixels and narrow breaks with a length of less than 1 pixel, and the boundaries are obviously jagged; the right side shows the result after the 5×5 rhombus structure element closing operation, the holes are completely filled, the breaks are repaired, and the jagged edges of the boundaries are eliminated, which intuitively demonstrates the effect of the peak period optimization parameters on the improvement of boundary continuity.

[0045] If the measured tide level at a certain station is ≤ 2.0 m, the area affected by that station is determined to be in low tide. During low tide, a large area of ​​tidal flats is exposed, which is prone to isolated noise points (such as residual small water bodies or cloud shadows misjudged as water bodies) and small non-water areas (such as exposed mudflats in tidal channels). At the same time, fine boundaries such as tidal channels need to be preserved. Therefore, for the preliminary water mask image part corresponding to this area, a 3×3 square structural element is used for opening operation (erosion followed by expansion) to eliminate isolated noise points and small non-water areas, while enhancing fine boundary details. Figure 6 This section showcases a comparison of morphological treatment details during the low ebb phase (for ease of visual comparison). Figure 6 (The raster mask has been converted into a vector polygon): The left side shows the result before processing, with multiple isolated noise points, blurred boundaries of tidal channel branches, and some small tidal channels interrupted due to noise interference; the right side shows the result after the 3×3 square structural element opening operation, where noise points are completely eliminated, interrupted tidal channel branches are reconnected, and the boundaries clearly show the distribution characteristics of tidal channels, verifying the effectiveness of the low tide period optimization parameters in preserving details.

[0046] After applying the above-mentioned tidal-driven morphological optimization operation, the optimized water mask map of the entire study area was obtained.

[0047] (5) Coastline extraction and accuracy verification: Coastline extraction: a) Raster to Vector: Convert the optimized water mask (raster) obtained in step (4) into a vector polygon layer. b) Boundary line extraction: Extract the outer boundary line (Edge) of the above water body polygon layer. This boundary line is the preliminary coastline.

[0048] Coastline post-processing: a) Fragment removal: Set the minimum polygon area threshold to 300 square meters, and delete tiny fragmented polygons with an area smaller than this threshold and their corresponding coastline fragments. b) Smoothing of linear features: The initial coastline is smoothed by applying the B-spline curve fitting algorithm. The node tolerance parameter is set to 10 meters to reduce jagged edges and generate a smoother curve that is more in line with the natural shape. c) Topology check and repair: Perform topology rule checks (ensure no self-intersections, no dangling nodes, and continuity without breaks), and automatically repair any topology errors found (such as node snapping, closing small gaps, and pruning self-intersection segments) to ensure that a topologically consistent and continuous coastline is generated.

[0049] The final result is high-precision, continuous, and smooth coastline vector data for the target area.

[0050] To verify the effectiveness of this method, this embodiment uses the coastline generated by visual interpretation of Gaofen-7 satellite images with a resolution better than 1 meter from the same period in the study area as the reference ground truth. Figure 7 The image shows the accuracy verification of coastline extraction based on high-resolution imagery. The red solid line in the image is the reference coastline interpreted by Gaofen-7, and the blue solid line is the coastline extracted by the method of this invention. The background is Gaofen-7 imagery. From the perspective of spatial matching, the two coastlines highly overlap in most areas, with only slight deviations in some silty tidal flats and dense tidal channels. The deviation values ​​are all <8 meters, which is consistent with the RMSE results in Table 2. This embodiment evaluates the location accuracy and morphological similarity by calculating the root mean square error (RMSE) and Hausdorff distance (HD) between the final coastline extracted by this method and the reference ground value. The comparison method is the traditional method of using NDWI + Otsu segmentation + fixed morphology (5×5 square closing operation). The specific data of the comparison experiment are shown in Table 2. Table 2 Comparison of Coastline Extraction Accuracy at Sites in the Study Area

[0051] As shown in Table 2, the RMSE of this method is reduced by an average of about 25% and HD by an average of about 18% in silty tidal flat areas (such as near stations A1 and A8). The accuracy is also steadily improved in sandy and bedrock coastline areas. In particular, in the boundary ambiguity area during high tide (such as near station A10) and the boundary fracture area during low tide (such as near station A8), this method significantly improves the continuity and detail preservation of the coastline, proving the effectiveness of adaptive decision-making of environmental parameters and tidal-driven optimization.

[0052] Example 3 like Figure 8 As shown, this embodiment provides a coastline intelligent extraction system based on adaptive environmental parameters, applying the method described in embodiment 1 or 2, including: Data acquisition module 301: used to acquire remote sensing image data of the target area of ​​the coastline to be extracted and the tidal water level data of the imaging time phase, and to preprocess the remote sensing image data. Water index calculation module 302: used to invert and calculate the suspended sediment concentration (SSC) of the target area based on the preprocessed remote sensing image data, and dynamically select the water index according to the suspended sediment concentration (SSC) to calculate the adaptive water index map corresponding to the target area. Preliminary segmentation module 303: used to perform threshold segmentation on the adaptive water index map to obtain a preliminary water mask map; Boundary optimization module 304: is used to determine the tidal period of the imaging phase based on the tidal water level data, and to perform boundary optimization on the preliminary water body mask map based on the dynamic selection of morphological operations according to the tidal period, so as to obtain the optimized water body mask map. Coastline extraction module 305: used to extract the coastline from the optimized water mask map and perform post-processing to obtain the final target area coastline.

[0053] In the specific implementation process, the data acquisition module 301 first acquires the remote sensing image data of the target area of ​​the coastline to be extracted and the tidal water level data of the imaging time phase, and preprocesses the remote sensing image data. Secondly, the water index calculation module 302 calculates the suspended sediment concentration (SSC) of the target area based on the preprocessed remote sensing image data, and dynamically selects the water index according to the suspended sediment concentration (SSC) to calculate the adaptive water index map corresponding to the target area. Next, the preliminary segmentation module 303 performs threshold segmentation on the adaptive water index map to obtain a preliminary water mask map; Then, the boundary optimization module 304 determines the tidal period of the imaging phase based on the tidal water level data, and performs boundary optimization on the preliminary water mask map by dynamically selecting morphological operations based on the tidal period, thereby obtaining the optimized water mask map. Finally, the coastline extraction module 305 extracts the coastline from the optimized water mask map and performs post-processing to obtain the final target area coastline.

[0054] In summary, this invention addresses the problem of traditional water body indices failing due to large differences in nearshore suspended sediment concentrations. It innovatively proposes an adaptive index selection strategy based on environmental parameters (suspended sediment concentration and tidal phase). By establishing a "sediment concentration-index performance" mapping matrix, the optimal index combination is dynamically selected in real time, effectively improving the accuracy of water body classification in complex environments. Furthermore, this invention addresses the characteristic of tidal flat boundaries undergoing morphological changes under tidal action by proposing a dynamic response mechanism between tidal level and morphological processing. By establishing a precise mapping relationship between tidal height and structural element type, rhomboid structural elements are used for closed-loop operations to suppress noise during high tide, while square structural elements are used for open-loop operations to enhance details during low tide. This effectively solves the boundary breakage problem caused by traditional fixed-parameter morphological processing, improving boundary continuity by approximately 25% in silty tidal flat areas.

[0055] The same or similar labels correspond to the same or similar parts; The terms used to describe positional relationships in the accompanying drawings are for illustrative purposes only and should not be construed as limiting this application. Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the implementation of the present invention. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively describe all embodiments here. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the claims of the present invention.

Claims

1. A method for intelligent extraction of coastlines based on adaptive environmental parameters, characterized in that, It includes the following steps: S1: Obtain the remote sensing image data of the target area where the coastline is to be extracted and the tidal water level data at the imaging time, and preprocess the remote sensing image data; S2: Invert and calculate the suspended sediment concentration SSC of the target area based on the preprocessed remote sensing image data, dynamically select a water body index according to the suspended sediment concentration SSC, and calculate the corresponding adaptive water body index map of the target area; S3: Perform threshold segmentation on the adaptive water body index map to obtain a preliminary water body mask map; S4: According to the tidal water level data, judge the tidal period at the imaging time, and dynamically select morphological operations based on the tidal period to optimize the boundary of the preliminary water body mask map to obtain an optimized water body mask map; S5: Extract the coastline from the optimized water body mask map and perform post-processing to obtain the final coastline of the target area.

2. The intelligent coastline extraction method based on adaptive environmental parameters according to claim 1, characterized in that, In step S1, the remote sensing image data is specifically multispectral remote sensing image data, and the spectral bands at least include the blue light band, the green light band, the red light band, the near-infrared band, and the short-wave infrared band; The preprocessing at least includes any one or more of radiometric calibration, atmospheric correction, geometric precise correction, and cloud masking.

3. The intelligent coastline extraction method based on adaptive environmental parameters according to claim 1, characterized in that, In step S2, calculate the suspended sediment concentration SSC of the target area based on the water color remote sensing inversion model, expressed as: in, The remote sensing reflectance is obtained from the preprocessed remote sensing image data; and These are the red and green wavelengths corresponding to the preprocessed remote sensing image data, respectively. , and These are the first to third empirical parameters, respectively.

4. The intelligent coastline extraction method based on adaptive environmental parameters according to claim 1, characterized in that, In step S2, dynamically select a water body index according to the suspended sediment concentration SSC, including: When SSC ≤ the first preset threshold, select the NDWI water body index; When the first preset threshold < SSC ≤ the second preset threshold, select the MNDWI water body index; When SSC > the second preset threshold, select a fused water body index, where the fused water body index is a linear weighted fusion of the IWI water body index and the NDWI water body index.

5. The intelligent coastline extraction method based on adaptive environmental parameters according to claim 1, characterized in that, In step S3, use the Otsu threshold segmentation algorithm to perform binary segmentation of water body and non-water body on the adaptive water body index map to obtain a preliminary water body mask map.

6. The intelligent coastline extraction method based on adaptive environmental parameters according to claim 1, characterized in that, In step S4, when the tidal water level is greater than the third preset threshold, it is determined as the high tide period, and closing operation is used for boundary optimization; When the tidal water level is less than or equal to the third preset threshold, it is determined as the low tide period, and opening operation is used for boundary optimization.

7. The intelligent coastline extraction method based on adaptive environmental parameters according to claim 6, characterized in that, The closing operation uses a diamond-shaped structuring element for boundary optimization, and the opening operation uses a square structuring element for boundary optimization.

8. The intelligent coastline extraction method based on adaptive environmental parameters according to claim 1, characterized in that, In step S5, convert the optimized water body mask map into a vector polygon layer, and use an edge detection algorithm to extract the water body boundary line in the vector polygon layer as the preliminary coastline; Perform post-processing on the preliminary coastline to obtain the final coastline of the target area.

9. The intelligent coastline extraction method based on adaptive environmental parameters according to claim 1, characterized in that, In step S5, the post-processing includes: speckle removal, smoothing processing based on the B-spline curve fitting algorithm, topological anomaly checking, and topological repair.

10. A coastline intelligent extraction system based on adaptive environmental parameters, employing the method described in any one of claims 1 to 9, characterized in that, It includes: Data acquisition module: used to obtain the remote sensing image data of the target area where the coastline is to be extracted and the tidal water level data at the imaging time, and preprocess the remote sensing image data; Water index calculation module: used to invert and calculate the suspended sediment concentration (SSC) of the target area based on the preprocessed remote sensing image data, and dynamically select the water index according to the suspended sediment concentration (SSC) to calculate the adaptive water index map corresponding to the target area. Preliminary segmentation module: used to perform threshold segmentation on the adaptive water index map to obtain a preliminary water mask map; Boundary optimization module: used to determine the tidal period of the imaging phase based on the tidal water level data, and to perform boundary optimization on the preliminary water body mask map based on the dynamic selection of morphological operations according to the tidal period, so as to obtain the optimized water body mask map; Coastline extraction module: used to extract the coastline from the optimized water mask map and perform post-processing to obtain the final target area coastline.

Images