Tidal flat large-scale remote sensing intelligent extraction method

Abstract

The invention discloses a tidal flat large-scale remote sensing intelligent extraction method. The method comprises the following steps: extracting sample points; extracting shoreline data; determining a tidal flat extraction range based on shoreline data; constructing a remote sensing feature data set, calculating different features based on remote sensing data and synthesizing the different features into a new data set; construction of a classification decision tree: determining a proper segmentation threshold value by using distribution frequency distribution of sample points in the remote sensing feature data set, and constructing the decision tree; synthesizing the lowest tide level data, namely synthesizing the lowest tide level data by using the first 5% and 5-25% of images according to the proportion that each pixel in the three-year remote sensing data is identified as a tidal flat; and extracting a tidal flat space range, applying a classification decision tree on the synthesized lowest tide level data to obtain a product, and carrying out data post-processing and statistical mapping. Through remote sensing big data, large-scale minimum tide level data extraction and synthesis and tidal flat drawing are realized, and the robustness and the application range of the method are improved while the limitation of the existing method is overcome.

Description

Technical Field

[0001] The present invention relates to the field of remote sensing big data object recognition, and in particular to a large-scale remote sensing intelligent extraction method for tidal flats. Background Art

[0002] The coastal ecosystem is one of the most dynamic ecosystems in the world, with characteristics such as high spatiotemporal dynamics, interlaced land and sea influences, diverse ecosystems, and large external disturbances. As the most drastic natural geographical unit in the coastal zone, tidal flats are affected by multiple factors such as tidal dynamic coverage, vegetation invasion, and coastal reclamation. Tidal flat wetlands provide ecological service value for coastal protection and bird habitats. Mastering their high spatiotemporal dynamic change characteristics and understanding their spatial range provide scientific data support for coastal protection, management, restoration, and policy formulation.

[0003] The intertidal zone has brought challenges to the accurate mapping of tidal flats due to external factors such as diverse terrain conditions, random tidal coverage, and low accessibility. Traditional field sampling surveys are difficult to conduct large-scale surveys in tidal flat areas due to factors such as cost and accessibility. Thanks to the development of remote sensing technology and the growth of satellite data, large-scale tidal flat monitoring based on remote sensing big data has been carried out. Existing tidal flat mapping either adopts a surface cover classification strategy, which requires a lot of manual costs for data screening and sample training; or frequency segmentation based on big data, the tidal flat range is obtained by percentage threshold cutting, but the data at extremely low tide is ignored, resulting in data underestimation; or a scheme for constructing spectral index is adopted, but the classification accuracy is still not high and it cannot be used on a large scale. Therefore, in the context of the development of remote sensing big data, it is still difficult to realize large-scale intelligent extraction of tidal flats, and to accurately obtain the spatial distribution range of tidal flats at the lowest tide while satisfying the generalization and robustness of the method. Summary of the invention

[0004] The purpose of the present invention is to solve the shortcomings of current remote sensing technology in large-scale tidal flat intelligent extraction. By providing a large-scale remote sensing intelligent extraction method for tidal flats, a decision tree suitable for large-scale intelligent tidal flat extraction is constructed by constructing a minimum unit for tidal flat extraction and using a remote sensing feature data set. Furthermore, in the extraction results of long time series data, the tidal flat area in the minimum unit is sorted to achieve the synthesis of the lowest tide data and the extraction of the maximum tidal flat spatial range. The current situation of large data processing volume of the classification algorithm, the inability of the frequency method to obtain the lowest tidal flat data image, the interference of cloud and fog data, and the poor spatial generalization ability of specific indexes in traditional solutions is overcome, thereby improving the accuracy and efficiency of remote sensing big data in tidal flat intelligent extraction.

[0005] The specific technical solution for achieving the purpose of the present invention is: A large-scale remote sensing intelligent extraction method for tidal flats comprises the following steps: Step 1: Sample point extraction According to the publicly published literature on coastal tidal flat mapping, high-resolution low-tide satellite data better than one meter, and low-tide drone mapping data, two types of samples are obtained: tidal flat samples and non-tidal flat samples. Each type of sample contains the type of land feature, coordinate information, and sample acquisition time. The type of land feature within 100 meters of each sample point should be consistent, and the spatial geometric distance between different points should be greater than 150 meters. Step 2: Shoreline data extraction Through remote sensing data, artificial interactive drawing of shoreline data is carried out to separate land pixels and coastal pixels; the drawing principle is to ensure that one side of the shoreline is land elements including impervious surfaces, farmland and aquaculture ponds, and the other side is coastal elements including coastal salt marsh vegetation, mangroves, tidal flats and open waters; Step 3: Determine the spatial scope of tidal flat extraction Generate a buffer zone of 5 km on the land side and 10 km on the sea side using the shoreline data drawn in step 2, and generate several 20 km × 20 km grids that are closely connected and completely cover the buffer zone; each grid is regarded as the smallest unit for tidal flat extraction; Step 4: Build a remote sensing feature dataset According to the buffer zone determined in step 3, all remote sensing image surface reflectance products with cloud cover less than 50% in the three natural years covering the buffer zone are screened, and a matching remote sensing feature dataset is constructed for each surface reflectance product, including: (1) extracting the quality control bands of the remote sensing feature dataset, eliminating all pixels with cloud identification results of "high / medium / low confidence", and treating pixels with cloud identification results of "not detected" as available pixels; (2) extracting the data of each original band from the eliminated remote sensing feature dataset, and calculating a total of six features: NDVI, mNDWI, Max(R, NIR), DIFF(G, SWIRII), DIFF(NIR, G), and DIFF(NIR+R,G+B); (3) synthesizing the original band data and the calculated six features to form a remote sensing feature dataset; the calculation method of each feature dataset is as follows: 1. NDVI (Normalized Difference Vegetation Index): NDVI = (NIR - R) / (NIR + R); 2. mNDWI (Modified Normalized Difference Water Index): mNDWI = (G - SWIR1) / (G + SWIR1); 3. Max(R, NIR) (maximum value): Max(R, NIR) = max(R, NIR); 4. DIFF(G, SWIRII) (Difference 1): DIFF(G, SWIRII) = G - SWIRII; 5. DIFF(NIR, G) (Difference 2): DIFF(NIR, G) = NIR - G; 6. DIFF(NIR + R, G + B) (Difference 3): DIFF(NIR + R, G + B) = (NIR + R) - (G + B); Among them, G, R, NIR, SWIRI, SWIRII are the values ​​corresponding to the green band, red band, near infrared band, short-wave infrared first band and short-wave infrared second band respectively; Step 5: Classification decision tree construction According to the two types of samples containing the ground object type information and coordinate information obtained in step 1, the frequency distribution diagram of the two types of samples in each feature data set is obtained in the remote sensing feature data set constructed in step 4, the threshold is determined by recursive method, and the misclassification rate under different thresholds is calculated. The lowest misclassification rate is recorded as the optimal segmentation threshold; the classification decision tree for tidal flat pixel recognition is constructed, which includes the following three components: 1. Determination of high tide beach area NDVI > T1 ∧ Max(R, NIR) ≥ T4 ∧ DIFF(NIR, G) ≥ 0 ∧ DIFF(NIR + R,G + B) ≥ 0; 2. Determination of medium tidal flat area T1 ≥ NDVI > T2 ∧ mNDWI < T5 ∧ Max(R, NIR) ≥ T4 ∧ DIFF(NIR, G) ≥0 ∧ DIFF(NIR + R, G + B) ≥ 0; 3. Determination of low tidal flat area T2 ≥ NDVI > T3 ∧ mNDWI < T6 ∧ Max(R, NIR) ≥ T4 ∧ T7 > DIFF(NIR,G) ≥ T8 ∧ T9 > DIFF(NIR + R, G + B) ≥ 0; T1-T9 are the optimal segmentation thresholds for each feature in the remote sensing feature data set; all pixels identified as high tide beach area, medium tide beach area or low tide beach area are recorded as tide beach pixels; Step 6: Lowest tide data synthesis Using the classification decision tree obtained in step 5, extract tidal flat pixels from all remote sensing image products and their matching remote sensing feature data sets in step 4 to obtain tidal flat pixels and non-tidal flat pixels; in the overlapping area of ​​the tidal flat extraction minimum unit and the buffer zone in step 3, calculate the proportion of tidal flat pixels to all pixels, which is recorded as the tidal flat pixel ratio; The top 5% of remote sensing image products with the highest proportion of tidal flat pixels and their matching remote sensing feature data sets are extracted, and the lowest tide level synthetic data is obtained by averaging. Check whether the data has spatial missing data. If there is missing data, extract remote sensing image products with 5%-25% of tidal flat pixels and their matching remote sensing feature data sets, and calculate the average value to fill the missing part. The filled data is used as the lowest tide level data synthetic product. If there is no missing part, it is directly used as the lowest tide level data synthetic product. Step 7: Extraction of tidal flat spatial extent Based on the lowest tide data synthesized in step 6, the classification decision tree in step 5 is used to identify tidal flat pixels and obtain the tidal flat spatial distribution product; Step 8: Data post-processing and statistical mapping Set the projection for the tidal flat spatial distribution product obtained in step 7, merge the patches with an area of ​​less than 2 hectares into the nearest patch, calculate the total area of ​​the tidal flat and the area of ​​the tidal flat in each minimum extraction unit; use mapping software to perform symbolization processing to obtain a visualized tidal flat extraction product.

[0006] Compared with the prior art, the present invention has the following advantages: (1) With the support of remote sensing big data, the present invention uses sufficient time series data to achieve intelligent extraction of large-scale tidal flats.

[0007] (2) The present invention constructs the smallest unit for tidal flat extraction and extracts the lowest tidal flat, thereby avoiding the mapping errors caused by large tidal level differences in single-view images in traditional methods.

[0008] (3) The present invention improves the accuracy of traditional tidal flat classification by constructing a classification decision tree, further removes pixel interference such as clouds and mist, and synthesizes the lowest tide data based on sorting thinking, avoiding the phenomenon of extremely low tide being missed due to the frequency thinking adopted by traditional methods. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] Figure 1 is a flow chart of the present invention; Figure 2 Schematic diagram of a time-series Landsat 8 original image data set in an embodiment of the present invention; Figure 3 A schematic diagram showing the proportion of tidal flat pixels in all images of three years as a result of the classification decision tree extraction of an embodiment of the present invention; Figure 4 A schematic diagram of the lowest tide data in the Yangtze River Estuary region synthesized by an embodiment of the present invention; Figure 5 It is a schematic diagram of the extraction results of some minimum extraction units in the embodiments of the present invention. DETAILED DESCRIPTION

[0010] In order to make the technical scheme, purpose and advantages of the present invention more clearly expressed, the present invention is explained below in conjunction with the drawings in the embodiments of the present invention, so that ordinary practitioners in this profession can implement the present invention. The following embodiments are used as an example of the present invention and are only used for explanation. Obviously, without departing from the principle of the present invention, any non-creative modifications, improvements and equivalent schemes and other technical schemes that do not deviate from the scope of the present invention should be included in the protection scope of the present invention.

[0011] Example: See Figure 1 In this embodiment, a large-scale remote sensing intelligent extraction method for tidal flats is constructed by constructing the smallest unit for tidal flat extraction and using the remote sensing feature data set to construct a decision tree suitable for large-scale intelligent tidal flat extraction. Furthermore, in the extraction results of long time series data, the tidal flat area in the smallest unit is sorted to achieve the synthesis of the lowest tide data and the extraction of the largest tidal flat spatial range. The main steps are as follows: Step 1: Sample point extraction, including: obtaining the point information of tidal flat samples and non-tidal flat samples based on existing literature containing coastal tidal flats or vegetation maps, high-resolution low-tide satellite data better than one meter, low-tide drone mapping data, etc. The point should ensure that the type of land features within 100 meters is consistent, and the spatial geometric distance between different points should be greater than 150 meters.

[0012] Step 2: Extract shoreline data, including: drawing shoreline data manually and interactively through remote sensing data, separating land pixels and coastal pixels. The drawing principle is to ensure that one side of the shoreline is land elements (such as impervious surfaces, farmland, aquaculture ponds, etc.), and the other side is coastal elements (such as coastal salt marsh vegetation, mangroves, tidal flats, open waters, etc.).

[0013] Step 3: Determine the spatial scope of tidal flat extraction, including: generating a buffer zone of 5 km on the land side and 10 km on the ocean side through the shoreline data drawn in step 2, and generating several 20 km × 20 km grids that are closely connected and completely cover the above buffer zones. Each grid is regarded as the smallest unit for tidal flat extraction. In this embodiment, the spatial scope of tidal flat extraction is set in the Yangtze River Estuary area, starting from Nantong City, Jiangsu Province in the north and Jiaxing City, Zhejiang Province in the south.

[0014] Step 4: Construct a remote sensing feature dataset, including: based on the buffer zone determined in step 3, screen all remote sensing image surface reflectance products with cloud cover less than 50% in the three natural years covering the buffer zone, and construct a matching remote sensing feature dataset for each surface reflectance product. The construction method is as follows: (1) extract the quality control band of the remote sensing feature dataset, and remove the three types of pixels with high, medium and low cloud identification results; (2) extract the data of each original band from the removed remote sensing feature dataset, and calculate a total of six features: NDVI, mNDWI, Max(R, NIR), DIFF(G, SWIRII), DIFF(NIR, G), and DIFF(NIR+R,G+B); (3) synthesize the original band data and the calculated six features to form a remote sensing feature dataset. The calculation method of each feature dataset is as follows: 1. NDVI (Normalized Difference Vegetation Index): NDVI = (NIR - R) / (NIR + R); 2. mNDWI (Modified Normalized Difference Water Index): mNDWI = (G - SWIR1) / (G + SWIR1); 3. Max(R, NIR) (maximum value): Max(R, NIR) = max(R, NIR); 4. DIFF(G, SWIRII) (Difference 1): DIFF(G, SWIRII) = G – SWIRII; 5. DIFF(NIR, G) (Difference 2): DIFF(NIR, G) = NIR - G; 6. DIFF(NIR + R, G + B) (Difference 3): DIFF(NIR + R, G + B) = (NIR + R) - (G + B); Where G, R, NIR, SWIRI, SWIRII are the values ​​corresponding to the green band, red band, near infrared band, shortwave infrared first band, and shortwave infrared second band, respectively. In this embodiment, the time is set to 2019 to 2021, and the selected image is Landsat 8 OLI data. The schematic diagram of the original image data set is as follows: Figure 2Among them, G, R, NIR, SWIRI, and SWIRII correspond to the 3rd band (wavelength range of 0.525–0.600 µm), the 4th band (wavelength range of 0.630–0.680 µm), the 5th band (wavelength range of 0.845–0.885 µm), the 6th band (wavelength range of 1.560–1.660 µm), and the 7th band (wavelength range of 2.100–2.300 µm) of the Landsat 8 OLI data, respectively.

[0015] Step 5: Construction of classification decision tree, including: based on the tidal flat samples and non-tidal flat samples obtained in step 1, in the remote sensing feature data set constructed in step 4, obtain the frequency distribution diagram of the two types of samples in each feature data set, use the recursive method to determine the threshold, calculate the misclassification rate under different thresholds, and record the lowest misclassification rate as the optimal segmentation threshold. The construction of classification decision tree for tidal flat pixel identification includes three components: 1. Determination of high tide beach area NDVI ≥ T1 ∧ Max(R, NIR) ≥ T4 ∧ DIFF(NIR, G) ≥ 0 ∧ DIFF(NIR + R,G + B) ≥ 0; 2. Determination of medium tidal flat area T1 ≥ NDVI > T2 ∧ mNDWI < T5 ∧ Max(R, NIR) ≥ T4 ∧ DIFF(NIR, G) ≥0 ∧ DIFF(NIR + R, G + B) ≥ 0; 3. Determination of low tidal flat area T2 ≥ NDVI > T3 ∧ mNDWI < T6 ∧ Max(R, NIR) ≥ T4 ∧ T7 > DIFF(NIR,G) ≥ T8 ∧ T9 > DIFF(NIR + R, G + B) ≥ 0; T1-T9 are the optimal segmentation thresholds for each feature in the remote sensing feature dataset. All pixels in the high tide beach area, medium tide beach area and low tide beach area are recorded as tidal beach pixels.

[0016] In this embodiment, the thresholds T1-T9 are set to 0.1, 0.05, -0.2, 0.08, 0.5, 0.8, 0.03, -0.03, 0.1 according to the distribution of the two types of samples in step 1 and the characteristic data sets of Landsat 8 data in step 4.

[0017] Step 6: minimum tide data synthesis, including: using the classification decision tree obtained in step 5, extracting tidal flat pixels from all remote sensing image products and their matching remote sensing feature data sets in step 4 to obtain tidal flat pixels and non-tidal flat pixels. In the overlapping area of ​​the tidal flat extraction minimum unit and the buffer zone in step 3, calculate the proportion of tidal flat pixels to all pixels, recorded as the tidal flat pixel ratio Figure 3 . Figure 3 The classification decision tree extraction result of this embodiment is a pixel of a tidal flat. In all remote sensing images within three years, the proportion of pixels judged to be tidal flats is shown. The black and white transition colors represent the proportion from 0 to 1, that is, white represents 0% of the pixel being judged to be a tidal flat in all remote sensing images within three years, and black represents 100% of the pixel being judged to be a tidal flat in all remote sensing images within three years.

[0018] The top 5% of remote sensing image products with the highest proportion of tidal flat pixels and their matching remote sensing feature data sets are extracted, and the lowest tide level synthetic data is obtained by averaging. Check whether there are spatial gaps in the data. If there are gaps, extract remote sensing image products with 5%-25% of tidal flat pixels and their matching remote sensing feature data sets, and calculate the average value to fill the gaps. The filled data is used as the lowest tide level synthetic product; if there are no gaps, it is directly used as the lowest tide level synthetic product. In this embodiment, the synthesized lowest tide level data for the Yangtze River Estuary area is as follows: Figure 4 . Figure 4 Image A on the left is a schematic diagram of the top 5% of remote sensing image products with the highest proportion of tidal flat pixels. The white part is a schematic diagram of data-missing pixels, which are supplemented by remote sensing image products with tidal flat pixels accounting for 5%-25%, that is, Figure 4 The middle image, Image B, is the lowest tide data of the Yangtze River Estuary area synthesized from the remote sensing image products with the top 5% of tidal flat pixels and the remote sensing image products with tidal flat pixels accounting for 5%-25%, that is, Figure 4 Right image: Max Tidal Flats Image.

[0019] Step 7: Extract the spatial range of tidal flats. Based on the lowest tide data synthesized in step 6, the classification decision tree in step 5 is used to identify tidal flat pixels and obtain the tidal flat spatial distribution product. The schematic diagram of the extraction results of some minimum units is shown in the figure. Figure 5 . Figure 5 Several groups of grayscale images of remote sensing images cut into the smallest units and the corresponding binary images of tidal flat spatial distribution products are displayed in turn. The grayscale image is the real remote sensing image, the white part of the binary map is the area identified as tidal flat pixels, and the black part is the area identified as non-tidal flat pixels.

[0020] Step 8: Data post-processing and statistical mapping, set a suitable projection for the tidal flat spatial distribution product obtained in step 7, merge the patches with an area of ​​less than 2 hectares into the nearest patch, calculate the total area of ​​the tidal flat and the area of ​​the tidal flat in each minimum extraction unit. Use mapping software for symbolization processing to obtain a visual tidal flat extraction product.

[0021] Based on the above steps, the present invention can utilize remote sensing big data, adopt the concept of tidal flat minimum unit cutting, combine classification decision tree and land pixel sorting method, obtain minimum tide data, and realize intelligent extraction of tidal flat area in the largest range, overcoming the limitations of existing methods such as low classification accuracy, omission of extremely low tide levels, and poor generalization performance, and forming a set of efficient, simple, reproducible and robust remote sensing tidal flat intelligent extraction solutions.

[0022] The above is only a specific implementation of the present invention. Any modifications, improvements and equivalent solutions obtained without creative labor and other technical solutions that do not deviate from the scope of the present invention should fall within the protection scope of the present invention.

Claims

1. A large-scale remote sensing intelligent extraction method for tidal flats, characterized in that: The following steps are involved: Step 1: Sample point extraction Based on publicly published literature on coastal tidal flat mapping, high-resolution low-tide satellite data better than one meter, and low-tide drone mapping data, two types of samples were obtained: tidal flat samples and non-tidal flat samples. Each type of sample contains the type of feature, coordinate information, and sample acquisition time. The type of feature is consistent within 100 meters of each sample point, and the spatial geometric distance between different points is at least 150 meters; Step 2: Shoreline data extraction Through remote sensing data, artificial interactive drawing of shoreline data is performed to separate land pixels and coastal pixels; the drawing ensures that one side of the shoreline is land elements including impervious surfaces, farmland and aquaculture ponds, and the other side is coastal elements including coastal salt marsh vegetation, mangroves, tidal flats and open waters; Step 3: Determine the spatial scope of tidal flat extraction Generate a buffer zone of 5 km on the land side and 10 km on the sea side using the shoreline data drawn in step 2, and generate several 20 km × 20 km grids that are closely connected and completely cover the buffer zone; each grid is regarded as the smallest unit for tidal flat extraction; Step 4: Build a remote sensing feature dataset According to the buffer zone determined in step 3, all remote sensing image surface reflectance products with cloud cover less than 50% in the three natural years covering the buffer zone are screened, and a matching remote sensing feature dataset is constructed for each surface reflectance product, including: (1) extracting the quality control bands of the remote sensing feature dataset, eliminating all pixels with cloud identification results of "high / medium / low confidence", and treating pixels with cloud identification results of "not detected" as available pixels; (2) extracting the data of each original band from the eliminated remote sensing feature dataset, and calculating a total of six features: NDVI, mNDWI, Max(R, NIR), DIFF(G, SWIRII), DIFF(NIR,G), and DIFF(NIR+R,G+B); (3) synthesizing the original band data and the calculated six features to form a remote sensing feature dataset; the calculation method of each feature dataset is as follows:

1. NDVI (Normalized Difference Vegetation Index): NDVI = (NIR - R) / (NIR + R); 2. mNDWI (Modified Normalized Difference Water Index): mNDWI = (G - SWIR1) / (G + SWIR1); 3. Max(R, NIR) (maximum value): Max(R, NIR) = max(R, NIR); 4. DIFF(G, SWIRII) (Difference 1): DIFF(G, SWIRII) = G - SWIRII; 5. DIFF(NIR, G) (Difference 2): DIFF(NIR, G) = NIR - G; 6. DIFF(NIR + R, G + B) (Difference 3): DIFF(NIR + R, G + B) = (NIR + R) - (G + B); Among them, G, R, NIR, SWIRI, SWIRII are the values ​​corresponding to the green band, red band, near infrared band, short-wave infrared first band and short-wave infrared second band respectively; Step 5: Classification decision tree construction According to the two types of samples obtained in step 1, the frequency distribution diagram of the two types of samples in each feature data set is obtained in the remote sensing feature data set constructed in step 4. The threshold is determined by recursive method, and the misclassification rate under different thresholds is calculated. The lowest misclassification rate is recorded as the optimal segmentation threshold. The classification decision tree for tidal flat pixel recognition is constructed, which includes the following three components:

1. Determination of high tide beach area NDVI > T1 ∧ Max(R, NIR) ≥ T4 ∧ DIFF(NIR, G) ≥ 0 ∧ DIFF(NIR + R, G +B) ≥ 0; 2. Determination of medium tidal flat area T1 ≥ NDVI > T2 ∧ mNDWI < T5 ∧ Max(R, NIR) ≥ T4 ∧ DIFF(NIR, G) ≥ 0∧ DIFF(NIR + R, G + B) ≥ 0; 3. Determination of low tide beach area T2 ≥ NDVI > T3 ∧ mNDWI < T6 ∧ Max(R, NIR) ≥ T4 ∧ T7 > DIFF(NIR, G)≥ T8 ∧ T9 > DIFF(NIR + R, G + B) ≥ 0; T1-T9 are the optimal segmentation thresholds for each feature in the remote sensing feature data set; all pixels identified as high tide beach area, medium tide beach area or low tide beach area are recorded as tide beach pixels; Step 6: Lowest tide data synthesis Using the classification decision tree obtained in step 5, extract tidal flat pixels from all remote sensing image products and their matching remote sensing feature data sets in step 4 to obtain tidal flat pixels and non-tidal flat pixels; in the overlapping area of ​​the tidal flat extraction minimum unit and the buffer zone in step 3, calculate the proportion of tidal flat pixels to all pixels, which is recorded as the tidal flat pixel ratio; The top 5% of remote sensing image products with the highest proportion of tidal flat pixels and their matching remote sensing feature data sets are extracted, and the lowest tide level synthetic data is obtained by averaging. Check whether the data has spatial missing data. If there is missing data, extract remote sensing image products with 5%-25% of tidal flat pixels and their matching remote sensing feature data sets, and calculate the average value to fill the missing part. The filled data is used as the lowest tide level data synthetic product. If there is no missing part, it is directly used as the lowest tide level data synthetic product. Step 7: Extraction of tidal flat spatial extent Based on the lowest tide data synthesized in step 6, the classification decision tree in step 5 is used to identify tidal flat pixels and obtain the tidal flat spatial distribution product; Step 8: Data post-processing and statistical mapping Set the projection for the tidal flat spatial distribution product obtained in step 7, merge the patches with an area of ​​less than 2 hectares into the nearest patch, calculate the total area of ​​the tidal flat and the area of ​​the tidal flat in each minimum extraction unit; use mapping software to perform symbolization processing to obtain a visualized tidal flat extraction product.

Images