A coastal wetland classification method, device, equipment and storage medium

Abstract

The embodiments of the present invention provide a coastal wetland classification method, device, equipment, and storage medium, relating to the field of remote sensing image recognition technology. The coastal wetland classification method includes the following steps: S1. Obtaining a Landsat image set. S2. Combining high tide images, low tide images, summer images, and winter images based on the image set. S3. Obtaining a coarse classification range of mangroves based on the low tide images. S4. Obtaining a coarse classification range of salt marshes and a coarse classification range of tidal flats based on public data. S5. Extracting a valid remote sensing index set from the four images based on the coarse classification range of mangroves, the coarse classification range of salt marshes, and the coarse classification range of tidal flats. S6. Obtaining a set of remote sensing index thresholds for each wetland type in different images based on the valid remote sensing index set. S7. Obtaining a fine classification range of mangroves, a fine classification range of salt marshes, and a fine classification range of tidal flats based on the set of remote sensing index thresholds for each wetland type in different images, the high tide images, the low tide images, the summer images, and the winter images.

Description

Technical Field

[0001] The present invention relates to the technical field of remote sensing image recognition, and in particular to a coastal wetland classification method, device, equipment and storage medium. Background Art

[0002] Coastal wetlands, primarily composed of mangroves, salt marshes, and tidal flats, play a unique role in connecting terrestrial and marine ecosystems, providing important ecosystem services such as coastal protection, climate regulation, and carbon sequestration. Although they cover only 0.5% of the Earth's surface area, coastal wetlands harbor extremely high biodiversity and are one of the most productive ecosystems. At the same time, they are also among the most vulnerable ecosystems, vulnerable to irreversible changes under the severe threats of climate change and human interference. For example, sea level rise caused by climate change could alter the frequency and depth of floods, thereby changing the composition of coastal wetlands and, in turn, affecting the habitats of native species and carbon cycling processes.

[0003] With increasing coastal development intensity, coastal wetland areas have drastically decreased in size, and their landscape types and patterns have changed significantly. Therefore, identifying the spatial pattern of coastal wetlands is crucial for fully understanding the structural and functional changes in coastal blue carbon ecosystems.

[0004] Currently, dynamic monitoring of coastal wetlands is primarily achieved through remote sensing technology, which can identify the spatial patterns of coastal wetlands over large areas and long time series. However, due to the high degree of environmental variability in intertidal zones, remote sensing monitoring methods have limitations in capturing instantaneous events, making their application more challenging than monitoring terrestrial landscapes.

[0005] Previous studies have mostly used low-tide imagery to capture the locations of exposed coastal wetlands, manually selecting sample points to determine wetland classification thresholds. However, using only high- and low-tide imagery for coastal wetland classification based on tidal characteristics makes it difficult to effectively distinguish between mangroves and salt marshes, which exhibit similar inundation characteristics at extreme tides. Manual sample point selection is not only inefficient but also susceptible to subjective factors, resulting in poor stability and significant uncertainty in classification accuracy.

[0006] In view of this, the applicant filed this application after studying the existing technology. Summary of the Invention

[0007] The present invention provides a coastal wetland classification method, device, equipment and storage medium to improve at least one of the above technical problems.

[0008] In a first aspect, an embodiment of the present invention provides a coastal wetland classification method, which includes steps S1 to S7.

[0009] S1. Obtain a Landsat image collection with a duration of at least one year.

[0010] S2. Based on the Landsat image collection, intertidal zone high tide images, intertidal zone low tide images, intertidal zone summer images and intertidal zone winter images are generated.

[0011] S3. Based on the intertidal zone low tide image, the coarse classification range of mangroves is obtained using the enhanced mangrove vegetation index.

[0012] S4. Obtain the coarse classification range of salt marshes and tidal flats based on public data.

[0013] S5. Set multiple sample points based on the coarse classification ranges of mangroves, salt marshes, and tidal flats. Extract a set of valid remote sensing indices for the sample points from the intertidal zone high tide images, low tide images, summer images, and winter images based on the multiple sample points.

[0014] S6. Obtain a remote sensing index threshold set for each wetland type in different images based on the valid remote sensing index set.

[0015] S7. Obtain the mangrove subclassification range, salt marsh subclassification range, and tidal flat subclassification range based on the remote sensing index threshold sets of each wetland type in different images, the intertidal zone high tide images, the intertidal zone low tide images, the intertidal zone summer images, and the intertidal zone winter images.

[0016] In a second aspect, an embodiment of the present invention provides a coastal wetland classification device, comprising:

[0017] The image acquisition module is used to obtain a Landsat image set with a duration of no less than one year.

[0018] The image synthesis module is used to synthesize intertidal zone high tide images, intertidal zone low tide images, intertidal zone summer images and intertidal zone winter images based on Landsat images.

[0019] The first range acquisition module is used to obtain the coarse classification range of mangroves based on the intertidal zone low tide image and the enhanced mangrove vegetation index.

[0020] The second range acquisition module is used to obtain the coarse classification range of salt marshes and tidal flats based on public data.

[0021] The sample data extraction module is used to set multiple sample points based on the coarse classification ranges of mangroves, salt marshes, and tidal flats. Based on these multiple sample points, a set of valid remote sensing indices is extracted from intertidal zone high tide images, low tide images, summer images, and winter images.

[0022] The threshold extraction module is used to obtain the remote sensing index threshold set of each wetland type in different images based on the effective remote sensing index set.

[0023] The third range acquisition module is used to obtain the mangrove subclassification range, salt marsh subclassification range and tidal flat subclassification range based on the remote sensing index threshold set of each wetland type in different images, intertidal zone high tide images, intertidal zone low tide images, intertidal zone summer images and intertidal zone winter images.

[0024] In a third aspect, an embodiment of the present invention provides a coastal wetland classification device, comprising a processor, a memory, and a computer program stored in the memory. The computer program can be executed by the processor to implement a coastal wetland classification method as described in any paragraph of the first aspect.

[0025] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, characterized in that the computer-readable storage medium includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute a coastal wetland classification method as described in any paragraph of the first aspect.

[0026] By adopting the above technical solution, the present invention can achieve the following technical effects:

[0027] This embodiment of the present invention utilizes multi-temporal optical remote sensing imagery to automatically and randomly select sample points and set coastal wetland classification thresholds, eliminating the tedious process of manually selecting sampling points. Furthermore, by innovatively incorporating phenological and tidal characteristics as a basis for wetland classification, intertidal wetlands can be efficiently and accurately identified. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments. It should be understood that the following drawings only illustrate certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be obtained based on these drawings without paying any creative work.

[0029] Figure 1 It is a flowchart of the coastal wetland classification method.

[0030] Figure 2 It is a logical framework diagram of the coastal wetland classification method.

[0031] Figure 3 It is a schematic diagram for obtaining the remote sensing index threshold.

[0032] Figure 4 It is a schematic diagram for obtaining the intertidal zone.

[0033] Figure 5 This is a schematic diagram showing how to differentiate between mangroves and salt marshes using remote sensing indices.

[0034] Figure 6 This is a classification map of coastal wetlands.

[0035] Figure 7 It is a structural diagram of the coastal wetland classification device. DETAILED DESCRIPTION

[0036] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0037] Example 1

[0038] See also Figures 1 to 6 The first embodiment of the present invention provides a coastal wetland classification method, which can be performed by a coastal wetland classification device (hereinafter referred to as the coastal wetland classification device). In particular, the method is performed by one or more processors in the coastal wetland classification device to implement steps S1 to S7.

[0039] S1. Obtain a Landsat image collection with a duration of at least one year.

[0040] Specifically, a sequence of original Landsat images covering the target area with a duration of at least one year is obtained. These images are then subjected to radiometric correction, atmospheric correction, and cloud removal, resulting in a Landsat image set with a duration of at least one year.

[0041] It is understandable that the coastal wetland classification device can be an electronic device with computing capabilities, such as a portable notebook computer, a desktop computer, a server, a smart phone, or a tablet computer.

[0042] S2. Based on the Landsat image collection, intertidal zone high tide images, intertidal zone low tide images, intertidal zone summer images and intertidal zone winter images are generated.

[0043] Specifically, such as Figure 2 and Figure 3 As shown in the figure, a year's worth of Landsat imagery is assembled into four images: high tide, low tide, summer, and winter intertidal images. These images capture the tidal characteristics of the target area and the vegetation index phenological characteristics within the phenological characteristics. In subsequent steps, based on the acquisition range of these four images, we can more accurately determine the distribution of wetland types.

[0044] On the basis of the above embodiment, in an optional embodiment of the present invention, step S2 specifically includes steps S21 to S23.

[0045] S21. Synthesize an initial high tide image, an initial low tide image, an initial summer image, and an initial winter image based on the Landsat image set.

[0046] S22. Obtain the intertidal zone range according to the initial high tide image and the initial low tide image.

[0047] S23 , extracting intertidal zone high tide images, intertidal zone low tide images, intertidal zone summer images, and intertidal zone winter images from the initial high tide images, initial low tide images, initial summer images, and initial winter images according to the intertidal zone range.

[0048] Specifically, the present invention identifies the intertidal zone and obtains the distribution areas of mudflats, salt marshes and mangroves. Therefore, by obtaining the range of the intertidal zone, data of the target area is processed in a targeted manner to reduce the amount of data processing.

[0049] Furthermore, mangroves have a phenological characteristic of being evergreen all year round. Salt marshes have a phenological characteristic of growing in summer and withering in winter. In this embodiment, summer and winter images of the intertidal zone were extracted based on these phenological characteristics, providing a data foundation for accurately identifying the extent of mangroves and salt marshes.

[0050] Based on the above embodiment, in an optional embodiment of the present invention, step S21 specifically includes steps S211 to S214.

[0051] S211. Based on the Landsat image set, perform 90th percentile synthesis of the normalized vegetation index to obtain the initial low tide image.

[0052] S212. Based on the Landsat image set, perform maximum synthesis on the normalized difference water index to obtain an initial high tide image.

[0053] Specifically, such as Figure 4 As shown, Google Earth Engine was used to perform a 90th percentile composite of the Normalized Difference Vegetation Index (NDVI) band of each Landsat image in a one-year Landsat image collection to obtain a low tide image; and a quality mosaic was performed on the Normalized Difference Water Index (NDWI) of each Landsat image in a one-year Landsat image collection to obtain a high tide image.

[0054] S213. Based on the Landsat image collection, perform median synthesis on images from April to September to obtain an initial summer image.

[0055] S214. Based on the Landsat image collection, perform median synthesis on images from October to March to obtain an initial winter image.

[0056] Specifically, such as Figure 3 As shown, using the median operation in Google Earth Engine, all bands of the imagery from April to September are median-composite to obtain winter (declining season) images, and all bands of the imagery from October to March are median-composite to obtain summer (growing season) images.

[0057] On the basis of the above embodiment, in an optional embodiment of the present invention, step S22 specifically includes steps S221 to S225.

[0058] S221. Obtain a first seawater range having a normalized difference water index greater than 0.2 based on the initial high tide image.

[0059] S222. Acquire a second seawater range having elevation data less than 1 meter based on the first seawater range.

[0060] Specifically, in the high tide image obtained in step S2, the vector range with an NDWI greater than 0.2 is retained to represent the maximum seawater extent at high tide. Even after this processing, inland water pixel interference still exists. Because inland elevation data differs significantly from water bodies, this embodiment calculates the average elevation data (DEM) within the vector range and retains the range with an average DEM less than 1 to eliminate inland errors and obtain a more accurate maximum seawater extent.

[0061] S223. Obtain a third seawater range having a normalized difference water index less than 0 based on the initial low tide image.

[0062] S224. Acquire a fourth seawater range having a water body frequency index greater than 97.5 based on the third seawater range.

[0063] S225 . Acquire a fifth seawater range having elevation data less than 1 meter based on the fourth seawater range.

[0064] Specifically, in the low tide image obtained in step S2, the vector range with NDVI < 0 and water frequency index > 97.5 is retained to represent the minimum seawater range at low tide. Considering that there may still be interference from inland water pixels, the minimum seawater range still retains the range with average DEM < 1 to remove inland errors.

[0065] The water body frequency index refers to the ratio of the number of times each pixel in the Landsat image set in one year is judged to be a water body according to the normalized difference water body index to the total valid number of times.

[0066] S226. Obtain an intertidal zone range based on the second seawater range and the fifth seawater range.

[0067] Specifically, in the second seawater range and the fifth seawater range, the difference set of the second seawater range and the fifth seawater range (i.e., the intersection of the complement of the minimum seawater vector range and the maximum seawater vector range) is retained to reach the required intertidal zone range, such as Figure 4 shown.

[0068] S3. Based on the intertidal zone low tide image, the coarse classification range of mangroves is obtained using the Enhanced Mangrove Vegetation Index (EMVI).

[0069] Specifically, EMVI: (GREEN-SWIR2) / (SWIR-GREEN). The Enhanced Mangrove Vegetation Index (EMVI) uses the green band and two shortwave infrared bands to enhance the differences in greenness and canopy moisture between mangroves and other vegetation, and can effectively distinguish mangroves from other vegetation.

[0070] In an optional embodiment of the present invention, based on the above embodiment, step S3 specifically includes determining a threshold value of the Enhanced Mangrove Vegetation Index (EMVI) using the Otsu threshold method based on the intertidal zone low tide image. A coarse mangrove classification range is obtained based on the EMVI threshold value.

[0071] In the low tide image, both the salt marsh and the mangrove are exposed. In this embodiment, the Otsu threshold method is used to obtain the EMVI threshold to distinguish the mangrove area from other non-mangrove areas.

[0072] Specifically, based on the Google Earth Engine platform, the OTSU algorithm (Otsu threshold method) is used to determine the EMVI index threshold for low tide images in the intertidal zone. Each pixel in the low tide images of the intertidal zone is classified according to the EMVI index threshold, and the coarse classification data of mangroves in the area are automatically obtained.

[0073] S4. Obtaining the rough classification range of salt marshes and tidal flats based on public data. Preferably, step S4 specifically includes steps S41 to S42.

[0074] S41. Obtain the coarse classification range of salt marshes based on the salt marsh data in the public global 30-meter wetland data.

[0075] S42. Spatially overlay the publicly available global 30-meter wetland data and the publicly available global distribution data of tidal flat ecosystems to obtain the intersection range as the coarse classification range of the tidal flat.

[0076] Specifically, the salt marsh data from the 2020 Global 30-meter Wetland Data Product (GWL FCS30) were used as the coarse salt marsh classification data. At the same time, the publicly available global distribution data of tidal flat ecosystems in the target area was spatially overlaid with the 2020 Global 30-meter Wetland Data Product (GWL_FCS30) to obtain the intersection area as the calibrated coarse tidal flat classification data.

[0077] S5. Set multiple sample points based on the coarse classification ranges of mangroves, salt marshes, and tidal flats. Extract a set of valid remote sensing indices for the sample points from the intertidal zone high tide images, low tide images, summer images, and winter images based on the multiple sample points.

[0078] Specifically, existing studies have often used manual sampling to obtain thresholds. This method is time-consuming and labor-intensive, and has low randomness and accuracy. Therefore, the present invention uses a more convenient, stable, and more random automated method for extracting random sample points to improve the accuracy and convenience of long-term coastal wetland classification.

[0079] Based on the above embodiment, in an optional embodiment of the present invention, step S5 specifically includes steps S51 to S58.

[0080] S51. Repeat the cycle of extracting 100 different groups of random points based on the coarse classification ranges of mangroves, salt marshes, and tidal flats to obtain multiple tidal flat sample points, multiple mangrove sample points, and multiple salt marsh sample points, each group containing 100 sample points.

[0081] S52 : Extracting the NDVI index from the summer image of the intertidal zone according to the multiple mangrove sample points to obtain a first NDVI data set.

[0082] S53. Extracting the NDVI index from the winter image of the intertidal zone according to the multiple mangrove sample points to obtain a second NDVI data set.

[0083] S54 , extracting the NDVI index from the summer image of the intertidal zone according to the multiple salt marsh sample points to obtain a third NDVI data set.

[0084] S55 . Extracting the NDVI index from the intertidal zone winter image according to the multiple salt marsh sample points to obtain a fourth NDVI data set.

[0085] S56 , extracting the NDVI index and the NDWI index from the intertidal zone low tide image according to the plurality of tidal flat sample points, to obtain a fifth NDVI data set and a first NDWI data set.

[0086] S57 , extracting the NDWI index from the intertidal zone high tide image according to the plurality of tidal flat sample points to obtain a second NDWI data set.

[0087] S58. Calculate the mean and standard deviation of each data set respectively, retain the data within two standard deviations from the mean, and obtain the effective remote sensing index set of the sample point.

[0088] Specifically, the embodiment of the present invention uses the NDVI index and the NDWI index as judgment objects. In other embodiments, different indices can be extracted for different images, and the present invention does not make specific limitations on this.

[0089] Preferably, a large number of random points are extracted within the vector ranges (coarse classification ranges) of the mangroves, tidal flats, and salt marshes in each of the four images using the RandomPoint code in the Google Earth Engine (GEE) cloud computing platform. To improve the accuracy of the calculation results, this embodiment of the present invention uses a for loop to repeatedly select 100 different groups of random points for threshold calculation. Each group contains 100 random points.

[0090] According to the selected sample points, the specific band values ​​in the image are extracted to obtain the data set. Then the mean and standard deviation of each data set are calculated respectively, and the data within two standard deviations from the mean are retained as valid data to reduce the error, such as Figure 5 shown.

[0091] S6. Obtain a remote sensing index threshold set for each wetland type in different images based on the valid remote sensing index set.

[0092] like Figure 5 As shown in the figure, the normalized vegetation index of different wetland types has obvious differences. In order to accurately classify coastal wetlands, it is first necessary to compare the inundation characteristics of different wetland types as they change with tides, and use the NDWI and NDVI indices that can characterize tidal characteristics for classification. However, using only tidal characteristics cannot effectively distinguish mangroves and salt marshes with similar tidal characteristics. This invention innovatively introduces phenological characteristics (winter images, summer images) to distinguish evergreen mangroves from seasonal salt marshes.

[0093] Based on the above embodiment, in an optional embodiment of the present invention, step S6 specifically includes steps S61 to S69.

[0094] S61. According to the effective remote sensing index set, obtain the minimum NDVI value of the winter image of the intertidal zone within the coarse classification range of the mangroves as the first threshold.

[0095] S62. According to the effective remote sensing index set, obtain the maximum NDVI difference between the summer image and the winter image of the intertidal zone within the coarse classification range of the mangrove forest as the second threshold.

[0096] S63. Obtain the maximum NDVI value of the winter image of the intertidal zone within the coarse classification range of the salt marsh according to the effective remote sensing index set, as the third threshold.

[0097] S64. Obtain, based on the effective remote sensing index set, the minimum NDVI value of the winter image of the intertidal zone within the coarse classification range of the salt marsh as a fourth threshold.

[0098] S65. Obtain, based on the effective remote sensing index set, the maximum NDVI difference between the intertidal zone summer image and the intertidal zone winter image within the coarse classification range of the salt marsh as a fifth threshold.

[0099] S66. Obtain the maximum NDVI value of the intertidal zone low tide image within the coarse classification range of the tidal flat according to the effective remote sensing index set, as a sixth threshold.

[0100] S67. Obtain the maximum NDWI value of the intertidal zone low tide image within the coarse classification range of the tidal flat according to the effective remote sensing index set, as the seventh threshold.

[0101] S68. Obtain, based on the effective remote sensing index set, the minimum NDWI value of the intertidal zone high tide image within the coarse classification range of the tidal flat as an eighth threshold.

[0102] S69. Obtain a remote sensing index threshold set for each wetland type in different images based on the first threshold, the second threshold, the third threshold, the fourth threshold, the fifth threshold, the sixth threshold, the seventh threshold, and the eighth threshold.

[0103] Specifically, tidal flats refer to mudflats and sandy beaches that are exposed at low tide and submerged at high tide. Compared to mangroves and salt marshes, tidal flats are significantly affected by tidal fluctuations. Extracting the NDWI values ​​of tidal flat sample points can effectively represent tidal information. This embodiment of the present invention extracts the minimum NDWI value of tidal flat sample points in high tide images and the maximum NDWI value in low tide images as important indicators for distinguishing tidal flats from other wetlands and various inland land cover types.

[0104] Another key characteristic that distinguishes tidal flats from mangroves and salt marshes is that they are exposed at low tide and lack vegetation. Because NDVI is positively correlated with vegetation growth, bare tidal flats at low tide have a significantly lower NDVI than mangroves and salt marshes. Therefore, extracting and calculating the maximum NDVI value from low-tide images of tidal flat samples can effectively distinguish unvegetated tidal flats from other wetlands.

[0105] In addition to tidal flats, the intertidal zone also has evergreen mangroves and salt marshes whose greenness varies seasonally. Since the extreme tidal inundation characteristics of mangroves and salt marshes are similar, it is not possible to effectively distinguish these two types of coastal wetlands using only tidal characteristics. Therefore, the embodiment of the present invention introduces the key parameter of phenological characteristics, and performs median synthesis of images from January, February, March, October, November, and December each year to obtain winter images. Based on the weak seasonality of evergreen mangroves, seasonal salt marshes have the seasonal characteristics of growing in summer and declining in winter, while tidal flats have the characteristics of no vegetation cover in all seasons. According to the principle that NDVI is positively correlated with the growth status of green vegetation, the three types of wetlands can be effectively distinguished by automatically generating random sample points and calculating the minimum NDVI value of the mangrove sample points, the maximum NDVI value of the salt marsh sample points, and the minimum NDVI value of the salt marsh sample points in the winter image.

[0106] Evergreen mangroves have weak seasonality, while deciduous salt marshes have seasonal characteristics of growing in summer and declining in winter. The phenological differences of mangroves throughout the year are smaller than those of salt marshes. To further improve accuracy, the embodiment of the present invention uses the minimum difference between the NDVI of the summer and winter images of the salt marsh samples and the maximum difference between the NDVI of the summer and winter images of the mangrove samples as classification indicators to characterize phenological differences based on the above characteristics. The specific classification process is as follows: Figure 5 shown.

[0107] S7. Obtain the mangrove subclassification range, salt marsh subclassification range, and tidal flat subclassification range based on the remote sensing index threshold sets of each wetland type in different images, the intertidal zone high tide images, the intertidal zone low tide images, the intertidal zone summer images, and the intertidal zone winter images.

[0108] Based on the above embodiment, in an optional embodiment of the present invention, step S7 specifically includes steps S71 to S74.

[0109] S71. Obtain a summer-winter NDVI difference image based on the summer image and the winter image of the intertidal zone.

[0110] S72: Based on the remote sensing index thresholds for each wetland type in different images, the winter intertidal zone image, and the summer-winter NDVI difference image, obtain a first range in which the NDVI index in the winter intertidal zone image is greater than a first threshold, and a second range in which the NDVI difference in the summer-winter NDVI difference image is less than a second threshold. The mangrove subclassification range is obtained based on the intersection of the first and second ranges.

[0111] S73. Based on the remote sensing index threshold sets for each wetland type in different images, the intertidal zone winter image, and the summer-winter NDVI difference image, obtain a third range in the intertidal zone winter image where the NDVI index is less than a third threshold, a fourth range in the intertidal zone winter image where the NDVI index is greater than a fourth threshold, and a fifth range in the summer-winter NDVI difference image where the NDVI difference is greater than a fifth threshold. Obtain a salt marsh subclassification range based on the intersection of the third, fourth, and fifth ranges.

[0112] S74. Based on the remote sensing index threshold sets for each wetland type in different images, the intertidal zone high tide images, and the intertidal zone low tide images, obtain a sixth range in the intertidal zone low tide images where the NDVI index is less than a sixth threshold, a seventh range in the intertidal zone low tide images where the NDVI index is less than a seventh threshold, and an eighth range in the intertidal zone high tide images where the NDWI index is greater than an eighth threshold. The tidal flat subclassification range is obtained based on the intersection of the sixth, seventh, and eighth ranges.

[0113] Specifically, the Jiulong River Estuary Wetland is a large mudflat formed by years of sedimentation in the Jiulong River Basin. This area includes the Longhai Jiulong River Estuary Mangrove Forest Provincial Nature Reserve, located in the intertidal zone of the mudflats at the Jiulong River estuary. Taking the Jiulong River Estuary Wetland in Fujian Province as an example, after steps S1 to S6, the following is obtained:

[0114] The minimum NDVI index value of the mangrove sample point in the winter image is 0.32 (the first threshold);

[0115] The maximum value of the difference in NDVI between mangrove sample points in the summer and winter difference images is 0.11 (the second threshold);

[0116] The maximum NDVI index value of the salt marsh sample point in the winter image is 0.50 (the third threshold);

[0117] The minimum value of the NDVI index of the salt marsh sample point in the winter image is -0.06 (the fourth threshold);

[0118] The minimum value of the difference in NDVI between the salt marsh sample points in the summer and winter difference images is 0.23 (the fifth threshold);

[0119] The maximum value of the NDVI index of the tidal flat sample point in the low tide image is 0.2 (the sixth threshold);

[0120] The maximum value of the NDWI index of the tidal flat sample point in the low tide image is 0.31 (the seventh threshold);

[0121] The minimum value of the NDWI index of the tidal flat sample point in the high tide image is 0.42 (the eighth threshold).

[0122] like Figure 5As shown in the figure, according to the above threshold, the area with NDVI value > 0.32 (first threshold) in the winter image and the area with NDVI value < 0.11 (second threshold) in the summer-winter difference image (summer image minus winter image) are intersected. The intersection area is retained in the summer image of Jiulong River Estuary, which is the mangrove area.

[0123] like Figure 5 As shown in the figure, according to the above thresholds, the areas in the winter image that meet the conditions of 0.50 (third threshold) > NDVI > -0.06 (fourth threshold) and the areas in the summer-winter difference image where the NDVI value is > 0.23 (fifth threshold) are intersected, and the intersection area retained in the summer image of Jiulong River Estuary is the salt marsh area.

[0124] like Figure 5 As shown in the figure, according to the above thresholds, the area with NDVI < 0.42 (sixth threshold) in the low tide image, the area with NDWI value < 0.31 (seventh threshold) in the low tide image, and the area with NDWI value > 0.42 (eighth threshold) in the high tide image are intersected, and the intersection area retained in the low tide image of Jiulong River Estuary is the tidal flat area.

[0125] This embodiment of the present invention utilizes multi-temporal optical remote sensing imagery to automatically and randomly select sample points and set coastal wetland classification thresholds, eliminating the tedious process of manually selecting sampling points. Furthermore, by innovatively incorporating phenological and tidal characteristics as a basis for wetland classification, intertidal wetlands can be efficiently and accurately identified.

[0126] Based on the above embodiment, in an optional embodiment of the present invention, a coastal wetland classification method further includes step S8.

[0127] S8. Perform synthesis processing based on the mangrove subclassification range, salt marsh subclassification range, and tidal flat subclassification range to obtain a coastal wetland classification map.

[0128] Specifically, in the previous steps, mangroves, salt marshes and tidal flats are extracted respectively according to the threshold classification rules to obtain the interpreted coastal wetland classification map. In order to display it more intuitively, the embodiment of the present invention synthesizes the mangrove, salt marsh and tidal flat images extracted from steps S1 to S7 based on ArcGIS, thereby obtaining a large-scale, long-time series, high-precision coastal wetland interpretation image, such as Figure 6 shown.

[0129] Example 2

[0130] like Figure 7 As shown, an embodiment of the present invention provides a coastal wetland classification device, which includes:

[0131] Image acquisition module 1 is used to acquire a Landsat image set with a duration of no less than one year.

[0132] The image synthesis module 2 is used to synthesize intertidal zone high tide images, intertidal zone low tide images, intertidal zone summer images and intertidal zone winter images based on the Landsat image set.

[0133] The first range acquisition module 3 is used to obtain the coarse classification range of mangroves based on the intertidal zone low tide image and the enhanced mangrove vegetation index.

[0134] The second range acquisition module 4 is used to obtain the coarse classification range of the salt marsh and the coarse classification range of the tidal flat according to the public data.

[0135] The sample data extraction module 5 is configured to set a plurality of sample points based on the coarse classification ranges of mangroves, salt marshes, and tidal flats, respectively. The module then extracts a set of valid remote sensing indices for the sample points from the intertidal zone high tide images, low tide images, summer images, and winter images, based on the plurality of sample points.

[0136] The threshold extraction module 6 is used to obtain a remote sensing index threshold set of each wetland type in different images based on the effective remote sensing index set.

[0137] The third range acquisition module 7 is used to obtain the mangrove subclassification range, salt marsh subclassification range and tidal flat subclassification range based on the remote sensing index threshold set of each wetland type in different images, intertidal zone high tide images, intertidal zone low tide images, intertidal zone summer images and intertidal zone winter images.

[0138] Based on the above embodiment, in an optional embodiment of the present invention, the image synthesis module 2 specifically includes:

[0139] The initial image synthesis unit is used to synthesize an initial high tide image, an initial low tide image, an initial summer image and an initial winter image according to the Landsat image set.

[0140] The intertidal zone range acquisition unit is used to acquire the intertidal zone range according to the initial high tide image and the initial low tide image.

[0141] The intertidal zone image acquisition unit is used to extract the intertidal zone high tide image, the intertidal zone low tide image, the intertidal zone summer image and the intertidal zone winter image from the initial high tide image, the initial low tide image, the initial summer image and the initial winter image according to the intertidal zone range.

[0142] Based on the above embodiment, in an optional embodiment of the present invention, the initial image synthesis unit includes:

[0143] The initial low tide image acquisition subunit is used to perform 90th percentile synthesis of the normalized vegetation index based on the Landsat image set to obtain the initial low tide image.

[0144] The initial high tide image acquisition subunit is used to perform maximum synthesis of the normalized difference water index based on the Landsat image set to obtain the initial high tide image.

[0145] The initial summer image acquisition subunit is used to perform median value synthesis on images from April to September based on the Landsat image set to obtain the initial summer image.

[0146] The initial winter image acquisition subunit is used to perform median value synthesis on images from October to March based on the Landsat image set to obtain the initial winter image.

[0147] Based on the above embodiment, in an optional embodiment of the present invention, the intertidal zone range acquisition unit includes:

[0148] The first seawater range acquisition subunit is used to acquire a first seawater range with a normalized difference water index greater than 0.2 based on the initial high tide image.

[0149] The second seawater range acquisition subunit is used to acquire a second seawater range with elevation data less than 1 meter based on the first seawater range.

[0150] The third seawater range acquisition subunit is configured to acquire a third seawater range having a normalized difference water index less than 0 based on the initial low tide image.

[0151] The fourth seawater range acquisition subunit is configured to acquire, based on the third seawater range, a fourth seawater range having a water body frequency index greater than 97.5.

[0152] a fourth seawater range acquisition subunit, configured to acquire a fifth seawater range having elevation data less than 1 meter based on the fourth seawater range;

[0153] The intertidal zone range acquisition subunit is used to acquire the intertidal zone range according to the second seawater range and the fifth seawater range.

[0154] Based on the above embodiment, in an optional embodiment of the present invention, the first range acquisition module 3 is specifically configured to:

[0155] Based on the intertidal zone low tide image, the threshold of the Enhanced Mangrove Vegetation Index was determined using the Otsu threshold method. The coarse classification range of mangroves was obtained based on the threshold of the Enhanced Mangrove Vegetation Index.

[0156] Based on the above embodiment, in an optional embodiment of the present invention, the second range acquisition module 4 specifically includes:

[0157] The salt marsh coarse classification range acquisition unit is used to obtain the salt marsh coarse classification range based on the salt marsh data in the public global 30-meter wetland data.

[0158] The salt marsh coarse classification range acquisition unit is used to spatially overlay the public global 30-meter wetland data and the public global distribution data of tidal flat ecosystems to obtain the intersection range as the tidal flat coarse classification range.

[0159] Based on the above embodiment, in an optional embodiment of the present invention, the sample data extraction module 5 specifically includes:

[0160] The sample point acquisition unit is used to repeatedly extract 100 different groups of random points based on the coarse classification ranges of mangroves, salt marshes, and tidal flats, respectively, to obtain multiple tidal flat sample points, multiple mangrove sample points, and multiple salt marsh sample points, where each group contains 100 sample points.

[0161] The first data acquisition unit is used to extract the NDVI index from the intertidal zone summer image according to multiple mangrove sample points to acquire a first NDVI data set.

[0162] The second data acquisition unit is used to extract the NDVI index from the intertidal zone winter image according to the multiple mangrove sample points to acquire a second NDVI data set.

[0163] The third data acquisition unit is configured to extract the NDVI index from the intertidal zone summer image according to the plurality of salt marsh sample points to acquire a third NDVI data set.

[0164] The fourth data acquisition unit is used to extract the NDVI index from the intertidal zone winter image according to the multiple salt marsh sample points to acquire a fourth NDVI data set.

[0165] The fifth data acquisition unit is configured to extract the NDVI index and the NDWI index from the intertidal zone low tide image according to the plurality of tidal flat sample points, and acquire a fifth NDVI data set and the first NDWI data set.

[0166] The sixth data acquisition unit is configured to extract the NDWI index from the intertidal zone high tide image according to the plurality of tidal flat sample points, and acquire a second NDWI data set.

[0167] The data screening unit is used to calculate the mean and standard deviation of each data set, retain the data within two standard deviations from the mean, and obtain the effective remote sensing index set of the sample point.

[0168] Based on the above embodiment, in an optional embodiment of the present invention, the threshold extraction module 6 specifically includes:

[0169] The first threshold extraction unit is used to obtain the minimum NDVI value of the intertidal zone winter image within the coarse classification range of the mangroves according to the effective remote sensing index set, as the first threshold.

[0170] The second threshold extraction unit is used to obtain the maximum NDVI difference between the intertidal zone summer image and the intertidal zone winter image within the coarse classification range of the mangrove according to the effective remote sensing index set, as the second threshold.

[0171] The third threshold extraction unit is used to obtain the maximum NDVI value of the intertidal zone winter image within the coarse classification range of the salt marsh according to the effective remote sensing index set, as the third threshold.

[0172] The fourth threshold extraction unit is used to obtain the minimum NDVI value of the winter image of the intertidal zone within the coarse classification range of the salt marsh according to the effective remote sensing index set, as the fourth threshold.

[0173] The fifth threshold extraction unit is used to obtain the maximum NDVI difference between the intertidal zone summer image and the intertidal zone winter image within the coarse classification range of the salt marsh according to the effective remote sensing index set, as the fifth threshold.

[0174] The sixth threshold value extraction unit is used to obtain the maximum NDVI value of the intertidal zone low tide image within the coarse classification range of the tidal flat according to the effective remote sensing index set, as the sixth threshold value.

[0175] The seventh threshold extraction unit is used to obtain the maximum NDWI value of the intertidal zone low tide image within the coarse classification range of the tidal flat according to the effective remote sensing index set, as the seventh threshold value.

[0176] The eighth threshold extraction unit is used to obtain the minimum NDWI value of the intertidal zone high tide image within the coarse classification range of the tidal flat according to the effective remote sensing index set, as the eighth threshold value.

[0177] The remote sensing index threshold value set acquisition unit is used to obtain the remote sensing index threshold value set of each wetland type in different images based on the first threshold value, the second threshold value, the third threshold value, the fourth threshold value, the fifth threshold value, the sixth threshold value, the seventh threshold value and the eighth threshold value.

[0178] Based on the above embodiment, in an optional embodiment of the present invention, the third range acquisition module 7 specifically includes:

[0179] The difference unit is used to obtain the summer-winter NDVI difference image based on the intertidal zone summer image and the intertidal zone winter image.

[0180] The mangrove subclassification range acquisition unit is configured to obtain, based on a set of remote sensing index thresholds for each wetland type in different images, the intertidal zone winter image, and the summer-winter NDVI difference image, a first range within which the NDVI index in the intertidal zone winter image is greater than a first threshold, and a second range within which the NDVI difference in the summer-winter NDVI difference image is less than a second threshold. The mangrove subclassification range is then acquired based on the intersection of the first and second ranges.

[0181] The salt marsh subclassification range acquisition unit is configured to, based on a set of remote sensing index thresholds for each wetland type in different images, the intertidal zone winter image, and the summer-winter NDVI difference image, acquire a third range in the intertidal zone winter image where the NDVI index is less than a third threshold, a fourth range in the intertidal zone winter image where the NDVI index is greater than a fourth threshold, and a fifth range in the summer-winter NDVI difference image where the NDVI difference is greater than a fifth threshold. The salt marsh subclassification range is then acquired based on the intersection of the third, fourth, and fifth ranges.

[0182] The tidal flat subclassification range acquisition unit is configured to, based on the remote sensing index threshold sets for each wetland type in different images, intertidal high tide images, and intertidal low tide images, acquire a sixth range in the intertidal low tide images where the NDVI index is less than a sixth threshold, a seventh range in the intertidal low tide images where the NDWI index is less than a seventh threshold, and an eighth range in the intertidal high tide images where the NDWI index is greater than an eighth threshold. The tidal flat subclassification range is then acquired based on the intersection of the sixth, seventh, and eighth ranges.

[0183] Example 3:

[0184] An embodiment of the present invention provides a coastal wetland classification device, comprising a processor, a memory, and a computer program stored in the memory. The computer program can be executed by the processor to implement a coastal wetland classification method as described in any paragraph of the first embodiment.

[0185] Example 4:

[0186] An embodiment of the present invention provides a computer-readable storage medium, characterized in that the computer-readable storage medium includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute a coastal wetland classification method as described in any paragraph of Example 1.

[0187] In the several embodiments provided in the embodiments of the present invention, it should be understood that the disclosed devices and methods can also be implemented in other ways. The device and method embodiments described above are merely illustrative. For example, the flowcharts and block diagrams in the accompanying drawings show the possible architectures, functions, and operations of the devices, methods, and computer program products according to multiple embodiments of the present invention. In this regard, each box in the flowchart or block diagram can represent a module, program segment, or part of the code, which contains one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions marked in the boxes can also occur in an order different from that marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flowchart, as well as the combination of boxes in the block diagram and / or flowchart, can be implemented using a dedicated hardware-based system that performs the specified functions or actions, or can be implemented using a combination of dedicated hardware and computer instructions.

[0188] In addition, the functional modules in the various embodiments of the present invention may be integrated together to form an independent part, or each module may exist independently, or two or more modules may be integrated to form an independent part.

[0189] If the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, electronic device, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes various media that can store program code, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk. It should be noted that, in this article, the terms "include", "comprising" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device that includes a series of elements includes not only those elements, but also other elements that are not explicitly listed, or also includes elements inherent to such a process, method, article or device. Without further constraints, an element defined by the phrase "comprises a..." does not preclude the existence of additional identical elements in the process, method, article or apparatus that includes the element.

[0190] The terms used in the embodiments of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The singular forms "a", "an", "the" and "the" used in the embodiments of the present invention and the appended claims are also intended to include plural forms unless the context clearly indicates otherwise.

[0191] It should be understood that the term "and / or" as used herein is merely a description of the relationship between associated objects, indicating that three possible relationships exist. For example, "A and / or B" can represent: A exists alone, A and B exist simultaneously, or B exists alone. Furthermore, the character " / " in this document generally indicates that the associated objects are in an "or" relationship.

[0192] The word "if," as used herein, may be interpreted as "at the time of" or "when" or "in response to determining" or "in response to detecting," depending on the context. Similarly, the phrases "if it is determined" or "if (stated condition or event) is detected" may be interpreted as "when it is determined" or "in response to the determination" or "when detecting (stated condition or event)" or "in response to detecting (stated condition or event)," depending on the context.

[0193] The "first" and "second" mentioned in the embodiments are merely used to distinguish similar objects and do not represent a specific ordering of the objects. It is understood that the specific order or precedence of "first" and "second" can be interchanged where appropriate. It should be understood that the objects distinguished by "first" and "second" can be interchanged where appropriate, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein.

[0194] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. A coastal wetland classification method, characterized in that: Include: Obtain a Landsat image collection with a duration of at least one year; Landsat images are used to generate intertidal zone high tide images, intertidal zone low tide images, intertidal zone summer images, and intertidal zone winter images. Based on the intertidal zone low tide images, the enhanced mangrove vegetation index was used to obtain the coarse classification range of mangroves; Obtain the coarse classification range of salt marshes and tidal flats based on public data; A plurality of sample points are set according to the coarse classification range of mangroves, the coarse classification range of salt marshes, and the coarse classification range of tidal flats; and a set of effective remote sensing indices of the sample points is extracted from the intertidal zone high tide image, the intertidal zone low tide image, the intertidal zone summer image, and the intertidal zone winter image according to the plurality of sample points; Among them, according to the coarse classification range of mangroves, the coarse classification range of salt marshes, and the coarse classification range of tidal flats, 100 groups of different random points were repeatedly extracted to obtain multiple tidal flat sample points, multiple mangrove sample points, and multiple salt marsh sample points; Extracting NDVI index from the intertidal zone summer image and the intertidal zone winter image according to the plurality of mangrove sample points and the plurality of salt marsh sample points to obtain a first NDVI data set, a second NDVI data set, a third NDVI data set, and a fourth NDVI data set; Extracting NDVI and NDWI indices from the intertidal zone low tide image and extracting NDWI indices from the intertidal zone high tide image according to the plurality of tidal flat sample points, thereby obtaining a fifth NDVI data set, a first NDWI data set, and a second NDWI data set; Calculate the mean and standard deviation of each data set respectively, retain the data within two standard deviations from the mean, and obtain the effective remote sensing index set; According to the effective remote sensing index set, the remote sensing index threshold set of each wetland type in different images is obtained; Obtaining a mangrove subclassification range, a salt marsh subclassification range, and a tidal flat subclassification range based on a remote sensing index threshold set, the intertidal zone high tide image, the intertidal zone low tide image, the intertidal zone summer image, and the intertidal zone winter image; Among them, the summer and winter NDVI difference image is obtained based on the intertidal zone summer image and the intertidal zone winter image; Obtain a first range in the winter intertidal zone image where the NDVI index is greater than a first threshold, and a second range in the summer-winter NDVI difference image where the NDVI difference is less than a second threshold. The intersection of the first and second ranges is the mangrove subclassification range. Obtain the third range of the intertidal zone winter image where the NDVI index is less than the third threshold and the fourth range where the NDVI index is greater than the fourth threshold, and the fifth range of the summer-winter NDVI difference image where the NDVI difference is greater than the fifth threshold. The intersection of the third, fourth, and fifth ranges is the salt marsh subclassification range. The sixth range in which the NDVI index is less than the sixth threshold value and the seventh range in which the NDWI index is less than the seventh threshold value in the intertidal zone low tide image are obtained, as well as the eighth range in which the NDWI index is greater than the eighth threshold value in the intertidal zone high tide image are obtained. The intersection of the sixth range, the seventh range, and the eighth range is the tidal flat subclassification range.

2. The coastal wetland classification method according to claim 1, characterized in that: The Landsat image set is used to generate intertidal zone high tide images, intertidal zone low tide images, intertidal zone summer images, and intertidal zone winter images, specifically including: Generating an initial high tide image, an initial low tide image, an initial summer image, and an initial winter image according to the Landsat image set; Acquiring an intertidal zone range according to the initial high tide image and the initial low tide image; According to the intertidal zone range, an intertidal zone high tide image, an intertidal zone low tide image, an intertidal zone summer image and an intertidal zone winter image are extracted from the initial high tide image, the initial low tide image, the initial summer image and the initial winter image.

3. The coastal wetland classification method according to claim 2, characterized in that: Generating an initial high tide image, an initial low tide image, an initial summer image, and an initial winter image based on the Landsat image set, specifically comprising: Based on the Landsat image set, the normalized vegetation index is synthesized at 90 percentile to obtain an initial low tide image; Based on the Landsat image set, a maximum value synthesis of the normalized difference water index is performed to obtain an initial high tide image; Based on the Landsat image set, median values ​​of images from April to September are synthesized to obtain an initial summer image; Based on the Landsat image set, median composite images from October to March are performed to obtain an initial winter image; Acquiring an intertidal zone range according to the initial high tide image and the initial low tide image specifically includes: Acquiring a first seawater range having a normalized difference water index greater than 0.2 according to the initial high tide image; Acquire a second seawater range having elevation data less than 1 meter based on the first seawater range; Acquiring a third seawater range having a normalized difference water index less than 0 according to the initial low tide image; According to the third seawater range, obtaining a fourth seawater range having a water body frequency index greater than 97.5; Acquire a fifth seawater range having elevation data less than 1 meter based on the fourth seawater range; The intertidal zone range is acquired according to the second seawater range and the fifth seawater range.

4. The coastal wetland classification method according to claim 1, characterized in that: Based on the intertidal zone low tide image, the enhanced mangrove vegetation index is used to obtain the coarse classification range of mangroves, which specifically includes: Determining a threshold value of an enhanced mangrove vegetation index (EMVI) based on the intertidal zone low tide image using the Otsu threshold method; and obtaining a coarse classification range of the mangroves based on the threshold value of the EVI; The rough classification ranges of salt marshes and tidal flats are obtained based on public data, including: The coarse classification range of salt marshes was obtained based on the salt marsh data in the public global 30-meter wetland database; The publicly available global 30-meter wetland data and the publicly available global distribution data of tidal flat ecosystems were spatially superimposed to obtain the intersection range as the coarse classification range of tidal flats.

5. The coastal wetland classification method according to claim 1, characterized in that: According to the effective remote sensing index set, a remote sensing index threshold set of each wetland type in different images is obtained, specifically including: According to the effective remote sensing index set, the minimum NDVI value of the intertidal zone winter image within the coarse classification range of the mangroves is obtained as a first threshold; According to the effective remote sensing index set, a maximum NDVI difference between the intertidal zone summer image and the intertidal zone winter image within the coarse classification range of the mangroves is obtained as a second threshold; According to the effective remote sensing index set, obtaining the maximum NDVI value of the winter image of the intertidal zone within the coarse classification range of the salt marsh as a third threshold; According to the effective remote sensing index set, obtaining the minimum NDVI value of the winter image of the intertidal zone within the coarse classification range of the salt marsh as a fourth threshold; According to the effective remote sensing index set, a maximum value of the NDVI difference between the intertidal zone summer image and the intertidal zone winter image within the coarse classification range of the salt marsh is obtained as a fifth threshold; According to the effective remote sensing index set, obtaining the maximum NDVI value of the intertidal zone low tide image within the coarse classification range of the tidal flat as a sixth threshold; According to the effective remote sensing index set, the maximum NDWI value of the intertidal zone low tide image within the coarse classification range of the tidal flat is obtained as the seventh threshold; According to the effective remote sensing index set, obtaining the minimum NDWI value of the intertidal zone high tide image within the coarse classification range of the tidal flat as an eighth threshold; According to the first threshold, the second threshold, the third threshold, the fourth threshold, the fifth threshold, the sixth threshold, the seventh threshold and the eighth threshold, a set of remote sensing index thresholds for each wetland type in different images is obtained.

6. The coastal wetland classification method according to claim 1, characterized in that: Also includes: A coastal wetland classification map is obtained by performing synthesis processing based on the mangrove subclassification range, the salt marsh subclassification range, and the tidal flat subclassification range.

7. A coastal wetland classification device, characterized in that: Include: Image acquisition module, used to obtain Landsat image sets with a duration of no less than one year; An image synthesis module is used to generate an intertidal zone high tide image, an intertidal zone low tide image, an intertidal zone summer image, and an intertidal zone winter image based on the Landsat image set; A first range acquisition module is configured to acquire a coarse classification range of mangroves using an enhanced mangrove vegetation index based on the intertidal zone low tide image; The second range acquisition module is used to obtain the coarse classification range of salt marshes and tidal flats based on public data; The sample data extraction module is used to set a plurality of sample points according to the coarse classification range of the mangroves, the coarse classification range of the salt marsh and the coarse classification range of the tidal flat; and extract a set of effective remote sensing indexes of the sample points from the intertidal zone high tide image, the intertidal zone low tide image, the intertidal zone summer image and the intertidal zone winter image according to the plurality of sample points; wherein, according to the coarse classification range of the mangroves, the coarse classification range of the salt marsh and the coarse classification range of the tidal flat, 100 groups of different random points are repeatedly extracted to obtain a plurality of tidal flat sample points, a plurality of mangrove sample points and a plurality of salt marsh sample points; according to the plurality of mangrove sample points and a plurality of The salt marsh sample points extract the NDVI index from the intertidal zone summer images and the intertidal zone winter images to obtain the first NDVI data set, the second NDVI data set, the third NDVI data set, and the fourth NDVI data set; based on multiple tidal flat sample points, the NDVI index and the NDWI index are extracted from the intertidal zone low tide images, and the NDWI index is extracted from the intertidal zone high tide images to obtain the fifth NDVI data set, the first NDWI data set, and the second NDWI data set; the mean and standard deviation of each data set are calculated respectively, and the data within two standard deviations from the mean are retained to obtain the valid remote sensing index set; A threshold extraction module is used to obtain a remote sensing index threshold set of each wetland type in different images based on the effective remote sensing index set; The third range acquisition module is used to obtain the mangrove subclassification range, salt marsh subclassification range and tidal flat subclassification range according to the remote sensing index threshold set of each wetland type in different images, the intertidal zone high tide image, the intertidal zone low tide image, the intertidal zone summer image and the intertidal zone winter image; wherein, the summer-winter NDVI difference image is obtained according to the intertidal zone summer image and the intertidal zone winter image; a first range in which the NDVI index in the intertidal zone winter image is greater than the first threshold value, and a second range in which the NDVI difference in the summer-winter NDVI difference image is less than the second threshold value is obtained, and the intersection of the first range and the second range is the mangrove subclassification range. The subdivision range of the forest; obtain the third range in which the NDVI index is less than the third threshold value and the fourth range in which the NDVI index is greater than the fourth threshold value in the winter image of the intertidal zone, and the fifth range in which the NDVI difference is greater than the fifth threshold value in the summer and winter NDVI difference image. The intersection of the third range, the fourth range and the fifth range is the subdivision range of the salt marsh; obtain the sixth range in which the NDVI index is less than the sixth threshold value and the seventh range in which the NDWI index is less than the seventh threshold value in the low tide image of the intertidal zone, and the eighth range in which the NDWI index is greater than the eighth threshold value in the high tide image of the intertidal zone. The intersection of the sixth range, the seventh range and the eighth range is the subdivision range of the tidal flat.

8. A coastal wetland classification device, characterized in that: It comprises a processor, a memory, and a computer program stored in the memory; the computer program can be executed by the processor to implement a coastal wetland classification method as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute a coastal wetland classification method according to any one of claims 1 to 6.

Images