Desert remote sensing image lake extraction method based on SSI and water body index fusion

Abstract

The invention discloses a desert remote sensing image lake extraction method based on SSI and water body index fusion, and the method comprises the steps: calculating a suspended sediment index (SSI) and a water body index based on remote sensing image data obtained by a GEE platform; carrying out fusion processing on the suspended sediment index and the water body index to obtain a fusion image so as to enhance the distinction degree of the lake area and the background area; a classification method is provided through single-index threshold segmentation; and classifying the fused image based on the proposed classification method, and accurately extracting the lake area. The lake extraction accuracy in the desert environment can be effectively improved, especially in a lake area containing high-concentration silt and a desert area with high earth surface reflection. Through the fusion of the two technologies, the advantages of different spectral indexes can be fully utilized, and more accurate water body identification can be realized in a complex environment.

Description

Technical Field

[0001] The invention relates to a lake extraction method, and in particular to a lake extraction method for desert remote sensing images based on the fusion of SSI and water body index. Background Art

[0002] In the field of remote sensing image processing, lake extraction and identification is an important research direction in GIS, environmental monitoring and water resources management.

[0003] At present, most common lake extraction methods are based on water index (such as normalized difference water index NDWI, modified normalized difference water index MNDWI, etc.), which can achieve lake identification by enhancing the spectral characteristics of water bodies. However, due to the unique surface reflection characteristics and complex vegetation and water mixing problems in desert areas, conventional water index extraction methods perform poorly in this environment.

[0004] In desert areas, lakes often contain a large amount of suspended sediment, causing the water to exhibit spectral reflectance that is different from traditional water characteristics. Especially in water bodies with high suspended sediment content, traditional methods have difficulty distinguishing suspended sediment from the surrounding arid environment, resulting in blurred lake boundaries or extraction errors.

[0005] Therefore, an innovative method that can combine the spectral characteristics of suspended sediment and water bodies is needed to improve the accuracy of lake extraction in desert environments.

[0006] In view of this, the present invention is proposed. Summary of the invention

[0007] The purpose of the present invention is to provide a lake extraction method for desert remote sensing images based on the fusion of SSI and water body index, aiming to overcome the problems of insufficient accuracy and false detection faced by the prior art when extracting lakes in desert environments.

[0008] The objective of the present invention is achieved through the following technical solutions:

[0009] The method for extracting lakes from desert remote sensing images based on the fusion of SSI and water body index of the present invention comprises the following steps:

[0010] 1) Calculate the suspended sediment index and water index based on remote sensing image data obtained from the GEE platform;

[0011] The suspended sediment index is referred to as SSI, which refers to the proportion of the cumulative dry weight of sediment particles of various sizes suspended in a certain section of river to the total weight of sediment particles of all sizes. It is usually used to describe the suspended state and distribution characteristics of sediment in a river.

[0012] 2) The suspended sediment index and the water body index are fused to obtain a fused image to enhance the distinction between the lake area and the background area;

[0013] 3) Propose a classification method through single exponential threshold segmentation;

[0014] 4) Classify the fused image based on the proposed classification method and accurately extract the lake area.

[0015] Compared with the prior art, the lake extraction method of desert remote sensing images based on the fusion of SSI and water body index provided by the present invention can effectively improve the accuracy of lake extraction in desert environments, especially in lake areas with high concentrations of sediment and desert areas with strong surface reflection. Through the fusion of these two technologies, the advantages of different spectral indices can be fully utilized to achieve more accurate water body identification in complex environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A schematic diagram of the Google Earth Engine (GEE) platform interface and the suspended sediment index (SSI) calculation function construction process according to an embodiment of the present invention;

[0017] Figure 2 Schematic diagram of suspended sediment (SSI) data extraction in an embodiment of the present invention;

[0018] Figure 3 A schematic diagram of extracting water body data using a multi-index collaborative method in an embodiment of the present invention;

[0019] Figure 4 It is a schematic diagram of combining suspended sediment data with water body extraction data in an embodiment of the present invention;

[0020] Figure 5 It is a schematic diagram comparing the JRC water data (global water coverage dataset provided by the European Joint Research Center) with the details of the embodiment of the present invention.

[0021] in, Figure 1 The interface is a free technical disclosure of the applicant to the public and does not constitute a technical limitation of this application. The parts that are suspected to be unclear in the figure may be deemed as undisclosed. DETAILED DESCRIPTION

[0022] The following is a clear and complete description of the technical solutions in the embodiments of the present invention in conjunction with the drawings in the embodiments of the present invention; it is obvious that the described embodiments are only part of the embodiments of the present invention, not all of the embodiments, which does not constitute a limitation of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the protection scope of the present invention.

[0023] First, the terms that may be used in this article are explained as follows:

[0024] The term “and / or” means that either or both of them can be realized at the same time. For example, X and / or Y means both “X” or “Y” and “X and Y”.

[0025] The terms "include", "comprises", "contains", "has" or other descriptions with similar semantics should be interpreted as non-exclusive inclusion. For example, including certain technical feature elements (such as raw materials, components, ingredients, carriers, dosage forms, materials, dimensions, parts, components, mechanisms, devices, steps, procedures, methods, reaction conditions, processing conditions, parameters, algorithms, signals, data, products or products, etc.) should be interpreted as including not only certain technical feature elements explicitly listed, but also other technical feature elements known in the art that are not explicitly listed.

[0026] The term "consisting of..." means excluding any technical feature elements not explicitly listed. If this term is used in a claim, it will make the claim closed, so that it does not contain technical feature elements other than the technical feature elements explicitly listed, except for the conventional impurities related to them. If this term only appears in a clause of a claim, it only limits the elements explicitly listed in the clause, and the elements recorded in other clauses are not excluded from the overall claim.

[0027] Unless otherwise specified or limited, the terms "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense, for example: it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be an indirect connection through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in this article can be understood according to specific circumstances.

[0028] The orientation or position relationship indicated by terms such as "center", "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", "clockwise", "counterclockwise", etc. are based on the orientation or position relationship shown in the drawings and are only for the convenience and simplification of description, and do not explicitly or implicitly indicate that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation of this document.

[0029] The contents not described in detail in the examples of the present invention belong to the prior art known to professionals in the field. If no specific conditions are specified in the examples of the present invention, the conditions are carried out according to the conventional conditions in the field or the conditions recommended by the manufacturer. If the manufacturer is not specified in the reagents or instruments used in the examples of the present invention, they are all conventional products that can be purchased commercially.

[0030] The method for extracting lakes from desert remote sensing images based on the fusion of SSI and water body index of the present invention comprises the following steps:

[0031] 1) Calculate the suspended sediment index and water index based on remote sensing image data obtained from the GEE platform;

[0032] The suspended sediment index is referred to as SSI, which refers to the proportion of the cumulative dry weight of sediment particles of various sizes suspended in a certain section of river to the total weight of sediment particles of all sizes. It is usually used to describe the suspended state and distribution characteristics of sediment in a river.

[0033] 2) The suspended sediment index and the water body index are fused to obtain a fused image to enhance the distinction between the lake area and the background area;

[0034] 3) Propose a classification method through single exponential threshold segmentation;

[0035] 4) Classify the fused image based on the proposed classification method and accurately extract the lake area.

[0036] 2. According to claim 1, the method for extracting lakes from desert remote sensing images based on the fusion of SSI and water body index is characterized in that the remote sensing image data includes multispectral or hyperspectral image data, and the data source is a remote sensing satellite or unmanned aerial vehicle equipment.

[0037] The suspended sediment index is calculated by the formula.

[0038] The fusion process uses a weighted average method to balance the contributions of the suspended sediment index and the water body index in lake extraction.

[0039] The weights in the weighted average method are dynamically adjusted according to the reflection characteristics of the desert and the lake to improve the accuracy of lake extraction.

[0040] The water index is a normalized difference water index or a modified normalized difference water index;

[0041] MNDWI>NDVI or MNDWI>EVI, and EVI<0.1;

[0042] NDWI is the normalized difference water index;

[0043] MNDWI is the modified normalized difference water index;

[0044] NDVI is the Normalized Difference Vegetation Index;

[0045] EVI is the enhanced vegetation index;

[0046] According to the above relationship between NDVI and MNDWI, a water body extraction function is defined, and the water body frequency is calculated based on the results obtained by the water body extraction function:

[0047]

[0048] The water body extraction function determines whether a pixel belongs to a water body based on the following conditions:

[0049] The water body frequency indicates the proportion of water body pixels in the total pixels. Water body categories are divided according to the water body frequency:

[0050] Pixels with a water body frequency greater than or equal to 0.75 are classified as permanent water bodies, pixels with a water body frequency less than or equal to 0.25 are classified as non-water bodies, and pixels with a water body frequency greater than 0.25 and less than 0.75 are classified as seasonal water bodies.

[0051] The suspended sediment index is calculated by the following formula:

[0052]

[0053] in:

[0054] R NIR is the reflectivity in the near-infrared band;

[0055] R RED is the reflectivity of the red band;

[0056] The index is used to identify the concentration of suspended sediment in lakes and helps distinguish between water areas with high sediment content and non-water areas on the desert surface.

[0057] The normalized difference water index was calculated by the following formula:

[0058]

[0059] in:

[0060] R NIR is the reflectivity in the near-infrared band

[0061] R RREEN is the reflectivity of the green band

[0062] The full name of this index is Normalized Difference Water Index. It uses normalized difference processing on specific bands of remote sensing images to highlight the water information in the image. This index enhances the contrast between water and vegetation and is suitable for water extraction, especially in areas with dense vegetation.

[0063] The modified normalized difference water index was calculated by the following formula:

[0064]

[0065] in:

[0066] R SWIR is the reflectivity in the short-wave infrared band

[0067] R GREEN is the reflectivity of the green band

[0068] The full name of the index is Modified Normalized Difference Water Index. It is calculated by using the reflectance values ​​of the green light band GREEN and the shortwave infrared band SWIR in remote sensing images to more accurately reflect the water body information. The index enhances the contrast between water bodies and exposed surfaces and is suitable for complex areas with buildings or desert surfaces.

[0069] The Enhanced Vegetation Index is calculated using the following formula:

[0070]

[0071] in:

[0072] R NIR is the reflectivity in the near-infrared band;

[0073] R RED is the reflectivity of the red band;

[0074] R BLUE is the reflectivity of the blue band;

[0075] G is the gain factor, set to 2.5;

[0076] C1 and C2 are coefficients for atmospheric correction, with values ​​of 6 and 7.5;

[0077] L is the influence of soil background, which is set to 1;

[0078] The full name of the index is Enhanced Vegetation Index. It is a vegetation index obtained by combining data from multiple remote sensing bands, especially the red band, near-infrared band and blue band, through specific mathematical calculations. It aims to overcome the shortcomings of the normalized vegetation index, which is prone to saturation in some cases and is greatly affected by soil moisture, so as to more accurately reflect the growth conditions and coverage of vegetation.

[0079] In summary, the lake extraction method for desert remote sensing images based on the fusion of suspended sediment index (SSI) and water index in the embodiment of the present invention analyzes remote sensing images. This method uses a variety of water indexes and comprehensively utilizes the spectral differences between water bodies and surface materials to significantly improve the accuracy of lake identification. Experimental results show that this method has better recognition performance than traditional methods in extracting lakes in desert areas, especially in lakes with high sediment content. It is particularly suitable for lake identification in desert environments.

[0080] The present invention can effectively distinguish between silt-containing lakes and the surrounding desert vegetation and surface, is suitable for extracting lakes from complex desert remote sensing images, and can significantly improve the accuracy of lake identification, especially for lake areas with high silt content, solving the problem of unsatisfactory identification effect of prior art in desert environments.

[0081] The present invention can combine the unique surface features of the desert with the spectral information of water bodies, improve the accuracy and reliability of desert lake extraction, and effectively address technical problems in lake identification in desert areas.

[0082] Compared with the existing lake extraction technology, which usually relies on traditional water indexes such as the Normalized Difference Water Index (NDWI) and the Modified Normalized Difference Water Index (MNDWI), these methods are insufficient in extraction accuracy in desert areas due to the complex mixture of surface reflection characteristics and water bodies. In particular, water bodies with high suspended sediment concentrations have spectral characteristics similar to the surrounding arid environment, which makes the lake boundary blurred. The fusion technology of the present invention, which combines the suspended sediment index with the water index, can effectively distinguish between sediment and water in desert lakes and improve the extraction accuracy.

[0083] In order to more clearly demonstrate the technical solution and technical effects provided by the present invention, the embodiments of the present invention are described in detail with specific embodiments below.

[0084] The method for extracting lakes from desert remote sensing images based on the fusion of suspended sediment index (SSI) and water index of the present invention comprises extraction of suspended sediment index, extraction of water index and fusion of the two indexes;

[0085] The suspended sediment index (SSI) is a method for estimating the content of suspended sediment in water bodies by using the characteristics of the water body reflectance spectrum in remote sensing images. Suspended sediment has a great influence on the scattering characteristics of light, especially in the visible and near-infrared bands. The higher the sediment concentration in the water body, the greater its reflectivity. Through a specific band combination, the distribution of suspended sediment in the water body can be accurately extracted.

[0086] The water index fusion is to accurately identify water areas by combining multiple water indexes and using the spectral differences between water and other surface materials in remote sensing images. Common water indexes include NDWI (Normalized Difference Water Index) and MNDWI (Modified Normalized Difference Water Index). By fusing multiple water indexes, the accuracy of the extraction results can be improved, especially in complex environments (such as desert areas).

[0087] like Figures 1 to 5 As shown:

[0088] The specific implementation includes the following steps:

[0089] Including SSI, MNDVI, NDVI, EVI, etc.;

[0090] The suspended sediment index (SSI) is calculated using the following formula:

[0091]

[0092] in:

[0093] R NIR is the reflectivity in the near-infrared band

[0094] R RED is the reflectivity of the red band

[0095] The index is used to identify the concentration of suspended sediment in lakes and helps distinguish between water areas with high sediment content and non-water areas on the desert surface.

[0096] The normalized difference water index (NDWI) was calculated by the following formula:

[0097]

[0098] in:

[0099] R NIR is the reflectivity in the near-infrared band

[0100] R GREEN is the reflectivity of the green band

[0101] The index enhances the contrast between water and vegetation and is suitable for water extraction, especially in areas with dense vegetation.

[0102] The modified normalized difference water index (MNDWI) was calculated by the following formula:

[0103]

[0104] in:

[0105] R SWIR is the reflectivity in the short-wave infrared band

[0106] R RREEN is the reflectivity of the green band

[0107] This index enhances the contrast between water bodies and bare ground and is suitable for complex areas containing buildings or desert surfaces.

[0108] The Enhanced Vegetation Index (EVI) is calculated by the following formula:

[0109]

[0110] in:

[0111] R NIR is the reflectivity in the near-infrared band

[0112] R RED is the reflectivity of the red band

[0113] R BLUE is the reflectivity of the blue band

[0114] G is the gain factor (usually set to 2.5)

[0115] C1 and C2 are coefficients used for atmospheric correction (common values ​​are 6 and 7.5)

[0116] L is the influence of soil background (usually set to 1)

[0117] Based on the relationship between NDVI and mNDWI, a water body extraction function is defined. The water body extraction function determines whether a pixel belongs to a water body based on the following conditions:

[0118] MNDWI>EVI or mNDWI>NDVI

[0119] EVI<0.1

[0120] Based on the results obtained from the water body extraction function, the water body frequency is calculated. The water body frequency represents the proportion of water body pixels in the total pixels. Next, the water body categories are divided according to the water body frequency. Pixels with a water body frequency greater than or equal to 0.75 are classified as permanent water bodies, pixels with a water body frequency less than or equal to 0.25 are classified as non-water bodies, and pixels with a water body frequency greater than 0.25 and less than 0.75 are classified as seasonal water bodies. The calculation formula is:

[0121]

[0122] The present invention superimposes the SSI suspended sediment frequency distribution map with the frequency distribution map of the water body itself to form a comprehensive image. On this comprehensive image, a specific threshold standard is set to identify and screen out those areas with high sediment content and therefore judged as non-water bodies. In order to more accurately divide the boundary between water bodies and non-water bodies, we will also consider the weight factor of the water body boundary. When judging whether an area belongs to a water body, in addition to the level of sediment content, we will also consider the proximity of the area to the known water body boundary or other relevant features. The above steps can more accurately extract the actual boundaries of the water body, providing a reliable basis for subsequent analysis or application.

[0123] Specific implementation example (taking a desert area as an example):

[0124] 1. If Figure 1 As shown in the figure, the specific steps for obtaining the suspended sediment index (SSI) in Google Earth Engine (GEE) usually involve processing and analyzing satellite image data using the GEE platform. The following is an overview of the steps for obtaining SSI based on the GEE platform, which may need to be adjusted according to specific image data and requirements:

[0125] To obtain the suspended sediment index (SSI) on the GEE platform, you need to log in to the GEE platform and browse the datasets to select suitable Landsat data, such as the top of the atmosphere reflectance dataset of Landsat 8 or 9. Select suitable satellite images based on the study area, time range, and cloud cover, and perform necessary preprocessing, such as cloud removal, to improve data quality. Use GEE's JavaScript API to build the SSI calculation function, which is based on the SSI calculation formula and usually involves mathematical operations on different bands of the image. Apply the calculation function to the filtered image data to generate the SSI image. View the generated SSI image on the GEE map to ensure that it meets expectations, and you can choose to download the results locally. Throughout the process, you need to pay attention to data quality and accuracy, the correctness of the calculation formula, and the resource limitations of the GEE platform.

[0126] 2. If Figure 2 As shown, with the help of the GEE platform, satellite images of the study area were obtained, and relevant algorithms were used on the GEE platform to obtain the water volume rate data of a desert lake. The data was downloaded locally, and ArcGIS was used for deep processing to obtain the annual area data of a desert lake in 2023.

[0127] The multi-index synergy method is used to process the image to obtain relatively good quality water body extraction data. The relevant index relationship is as follows: (MNDWI>NDVI or MNDWI>EVI) and EVI<0.1

[0128] In the formula, MNDWI is the improved normalized water index, NDVI is the normalized vegetation index, and EVI (Enhanced Vegetation Index) is the enhanced vegetation index. This method is proposed mainly to eliminate the interference of vegetation on water extraction. Through the data obtained from actual operation, it is found that the lake area data extracted by this method in a certain desert is of better quality than traditional single indices such as MNDWI and DLWI.

[0129] Compared with the single index method, the multi-index collaborative method does not need to use the Otsu algorithm, and can directly use the size relationship between multiple indicators to extract water bodies, so there is no need for manual participation in outlining the boundaries of the lake area. In terms of image data, since there is no restriction on images by the Otsu algorithm, the multi-index collaborative method can use more images. After the image is screened, the preliminary work of lake water extraction is carried out on the GEE platform. The calculation formulas of MNDWI, NDNI, and EVI water indexes and the multi-index collaborative water extraction formula are constructed. For each screened image, water bodies are identified and extracted by the multi-index collaborative method, and the water body ratio is calculated. The normalized water index (NDWI) and the modified normalized water index (MNDWI) are calculated using the acquired image data. The calculation results of these two indices range from -1 to 1. Negative values ​​represent non-water features, and positive values ​​represent water features.

[0130] Based on the relationship between NDVI and MNDWI, a water body extraction function is defined. The water body extraction function determines whether a pixel belongs to a water body based on the following conditions:

[0131] mNDWI>EVI or mNDWI>NDVI EVI<0.1

[0132] Among them, EVI is the calculated enhanced vegetation index. Pixels that meet the above conditions are considered to be water bodies, otherwise they are considered to be non-water bodies.

[0133] 3. If Figure 3 , Figure 4As shown, the obtained SSI data is combined with the water body data extracted by multi-index collaborative extraction, and the threshold is set at 0.332. The solid sediment is judged as solid sediment, and the suspended sediment is judged as suspended sediment. Suspended sediment exists in water, so it is judged as water body, thereby improving the accuracy of water body extraction. The water body data is visually interpreted and each lake area is marked in order to obtain a more accurate threshold. The primary water body data is further processed in ArcGIS software. The raster-to-vector tool is used to vectorize the lake water body identification results, and the non-water body parts such as shadows are visually interpreted to obtain the lake water body vector data. The area calculation tool is used to obtain the area of ​​all lakes, and the lakes in a single picture are given corresponding names. The spatial connection tool is used to associate names with all lakes in the atlas.

[0134] 4. As shown in Tables 1, 2 and Figure 5 (a, b) and Figure 5 As shown in (c, b), data accuracy verification was performed. A confusion matrix was constructed, the Kappa coefficient was calculated, and the accuracy of the obtained lake area dataset was verified. The Google high-resolution image and the Jilin No. 1 dry-resolution image were visually interpreted. To ensure the accuracy of the verification data, non-water body verification points were selected within the 100m buffer zone of the lake shoreline, and water body verification points were randomly distributed inside the lake water body. A total of 200 verification points were obtained, and 100 Google high-resolution image verification points (100 non-water body verification points and 100 water body verification points) were obtained. The obtained verification points were checked through Google high-resolution historical images to ensure the accuracy of the verification point data.

[0135] (1) Single exponential threshold segmentation combined with Otsu algorithm uploads the verification point data to the GEE platform, and uses the "sampleRegions" function to count the attributes of the verification points in the extracted lake water data set. This function can add the attributes of the result data to the verification point data, thereby distinguishing the attributes of the verification points in the result data, counting the number of verification points, and then constructing a confusion matrix and calculating the Kappa coefficient. The results are as follows:

[0136] Table 1 JRC data water body data accuracy verification

[0137]

[0138] (2) The multi-index synergy method uses the verification point data and lake area data to construct a confusion matrix, calculate the Kappa coefficient, and verify the accuracy of a desert lake area dataset obtained by the multi-index synergy method. The results are as follows:

[0139] Table 2 Accuracy verification of water body dataset obtained by multi-index collaborative method

[0140]

[0141] Comparison of local details of the two water extraction results ( Figure 5 ), compared with the finished JRC water body data, the boundaries of the water bodies extracted using the hybrid index algorithm are more consistent with the boundaries of the water bodies in the actual synthetic remote sensing images. The JRC dataset has the problem that the extracted water area is smaller than the actual water area, which causes errors in water body extraction.

[0142] The above is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by any technician familiar with the technical field within the technical scope disclosed in the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims. The information disclosed in the background technology section of this article is only intended to deepen the understanding of the overall background technology of the present invention, and should not be regarded as an admission or in any form that the information constitutes prior art known to those skilled in the art.

Claims

1. A method for extracting lakes from desert remote sensing images based on the fusion of SSI and water index, characterized in that: The following steps are involved: 1) Calculate the suspended sediment index and water index based on remote sensing image data obtained from the GEE platform; The suspended sediment index is referred to as SSI, which refers to the proportion of the cumulative dry weight of sediment particles of various sizes suspended in a certain section of river to the total weight of sediment particles of all sizes. It is usually used to describe the suspended state and distribution characteristics of sediment in a river. 2) The suspended sediment index and the water body index are fused to obtain a fused image to enhance the distinction between the lake area and the background area; 3) Propose a classification method through single exponential threshold segmentation; 4) Classify the fused image based on the proposed classification method and accurately extract the lake area.

2. The method for extracting lakes from desert remote sensing images based on the fusion of SSI and water index according to claim 1 is characterized in that: The remote sensing image data includes multispectral or hyperspectral image data, and the data source is a remote sensing satellite or unmanned aerial vehicle equipment.

3. The method for extracting lakes from desert remote sensing images based on the fusion of SSI and water index according to claim 2 is characterized in that: The suspended sediment index is calculated by the formula.

4. The method for extracting lakes from desert remote sensing images based on the fusion of SSI and water index according to claim 3 is characterized in that: The fusion process uses a weighted average method to balance the contributions of the suspended sediment index and the water body index in lake extraction.

5. The method for extracting lakes from desert remote sensing images based on the fusion of SSI and water index according to claim 4 is characterized in that: The weights in the weighted average method are dynamically adjusted according to the reflection characteristics of the desert and the lake to improve the accuracy of lake extraction.

6. The method for extracting lakes from desert remote sensing images based on the fusion of SSI and water index according to any one of claims 1 to 5, characterized in that: The water index is a normalized difference water index or a modified normalized difference water index; MNDWI>NDVI or MNDWI>EVI, and EVI<0.1; NDWI is the normalized difference water index; MNDWI is the modified normalized difference water index; NDVI is the Normalized Difference Vegetation Index; EVI is the enhanced vegetation index; According to the above relationship between NDVI and MNDWI, a water body extraction function is defined, and the water body frequency is calculated based on the results obtained by the water body extraction function: The water body extraction function determines whether a pixel belongs to a water body based on the following conditions: The water body frequency indicates the proportion of water body pixels in the total pixels. Water body categories are divided according to the water body frequency: Pixels with a water body frequency greater than or equal to 0.75 are classified as permanent water bodies, pixels with a water body frequency less than or equal to 0.25 are classified as non-water bodies, and pixels with a water body frequency greater than 0.25 and less than 0.75 are classified as seasonal water bodies.

7. The method for extracting lakes from desert remote sensing images based on the fusion of SSI and water index according to claim 6, characterized in that: The suspended sediment index is calculated using the following formula: in: R NIR is the reflectivity in the near-infrared band; R RED is the reflectivity of the red band; The index is used to identify the concentration of suspended sediment in lakes and helps distinguish between water areas with high sediment content and non-water areas on the desert surface.

8. The method for extracting lakes from desert remote sensing images based on the fusion of SSI and water index according to claim 6, characterized in that: The normalized difference water index was calculated by the following formula: in: R NIR is the reflectivity in the near-infrared band R GREEN is the reflectivity of the green band The full name of this index is Normalized Difference Water Index. It uses normalized difference processing on specific bands of remote sensing images to highlight the water information in the image. This index enhances the contrast between water and vegetation and is suitable for water extraction, especially in areas with dense vegetation.

9. The method for extracting lakes from desert remote sensing images based on the fusion of SSI and water index according to claim 6, characterized in that: The modified normalized difference water index was calculated by the following formula: in: R SWIR is the reflectivity in the short-wave infrared band R GREEN is the reflectivity of the green band The full name of the index is Modified Normalized Difference Water Index. It is calculated by using the reflectance values ​​of the green light band GREEN and the shortwave infrared band SWIR in remote sensing images to more accurately reflect the water body information. The index enhances the contrast between water bodies and exposed surfaces and is suitable for complex areas with buildings or desert surfaces.

10. The method for extracting lakes from desert remote sensing images based on the fusion of SSI and water index according to claim 6, characterized in that: The Enhanced Vegetation Index is calculated using the following formula: in: R NIR is the reflectivity in the near-infrared band; R RED is the reflectivity of the red band; R BLUE is the reflectivity of the blue band; G is the gain factor, set to 2.5; C1 and C2 are coefficients for atmospheric correction, with values ​​of 6 and 7.5; L is the influence of soil background, which is set to 1; The full name of the index is Enhanced Vegetation Index. It is a vegetation index obtained by combining data from multiple remote sensing bands, especially the red band, near-infrared band and blue band, through specific mathematical calculations. It aims to overcome the shortcomings of the normalized vegetation index, which is prone to saturation in some cases and is greatly affected by soil moisture, so as to more accurately reflect the growth conditions and coverage of vegetation.

Images