A method for identifying and estimating the area of a beach by remote sensing

Abstract

The present application relates to the field of ecology and remote sensing, and discloses a kind of beach identification and area remote sensing estimation method, comprising: determining ground index according to normalized vegetation index and normalized water index, and then obtaining pure water area;Based on the pure water area, the time when the maximum pure water area and the minimum pure water area correspond to the high and low water level line is determined;When obtaining the low water level line, the non-pure pixel water body proportion is obtained by using random forest method and microwave data, and then the low water level line is obtained;When obtaining the high water level line, the function of microwave data and water body proportion is established, and the non-pure pixel water body proportion is obtained, and then the high water level line is obtained;The beach type is obtained by using the high water level line and the low water level line combined with the beach identification model, and the beach area is determined. By using remote sensing technology to extract water body and land surface characteristics, through the extraction of high and low water level lines, combined with the distribution and morphological characteristics of beach, the precise identification of different beach types by remote sensing and the estimation of beach area are realized.

Description

Technical Field

[0001] This invention relates to the fields of ecology and remote sensing technology, and in particular to a method for identifying and estimating the area of ​​sandbars using remote sensing. Background Technology

[0002] A sandbar is a topographic feature that forms along the shoreline of a river or lake, or in the middle of a river. It is usually formed by the accumulation of sediments such as sand and silt, gradually forming with the movement of water. Sandbars are very important to the local ecological environment. The vegetation growing on sandbars provides food sources and habitats for some aquatic organisms, and the sediment deposited on them also provides fertile soil and sufficient arable land for agricultural production in some areas.

[0003] Shoals and sandbars are mainly distributed on the convex banks of lakes and rivers, and the land they form has become an indispensable part of the local ecological environment and human production and life. Based on location and shape, shoal and sandbar wetlands are mainly divided into river convex banks, lake shorelines, lake shoreline protrusions, and lake islands or islets. Different types of shoals and sandbars correspond to different development potentials and corresponding ecological protection measures. The microbial communities formed by shoals and lake islands on lake shorelines greatly enhance the biodiversity of the ecological environment and the self-purification capacity of the water body. Agricultural development on shoals and river convex banks provides crops with abundant inorganic salts through sediment deposition, and during the annual flood season, the soil is further deposited under floodwaters, increasing soil fertility. Therefore, to determine the development potential of shoals and implement different ecological protection measures for different types of shoals, we need to obtain specific distribution information for different types of shoals and sandbars.

[0004] Currently, methods for obtaining information on the distribution of river islands and shoals mainly include manual surveys and methods that first use remote sensing followed by manual screening. Manual surveys primarily involve on-site investigations to determine the distribution of river islands and shoals before calculating their area. The problem with this method is that it struggles to effectively and in real-time reflect spatiotemporal changes in research areas with significant differences. Remote sensing is a widely used method, characterized by its objectivity and convenience, and its ability to acquire information over large spatiotemporal spans. Many existing remote sensing methods exist, with a common one using time series analysis to extract areas of change between land and water bodies throughout the year to identify river islands and shoals. However, this method requires accurate high and low water levels, and currently, there are no particularly suitable models for accurately extracting high and low water levels over large spatial areas. Furthermore, it lacks the ability to accurately extract riverbanks, convex banks, and permanent water bodies in the center of river islands and shoals for identification.

[0005] To address the aforementioned issues, and considering the current limitations in accurately identifying high and low water levels over large areas and in accurately identifying river shoals over large regions, this method comprehensively considers methods for acquiring high and low water levels, the morphological characteristics of various river shoals, and their features in remote sensing images. It proposes a method for extracting river shoal distribution by acquiring high and low water levels and river morphological features through remote sensing, aiming to achieve simultaneous and high-precision extraction of river shoals over large areas. Summary of the Invention

[0006] The purpose of this invention is to provide a method for identifying and remotely estimating the area of ​​sandbars, aiming to solve the above-mentioned problems.

[0007] This invention provides a method for identifying and remotely estimating the area of ​​sandbars, including:

[0008] The surface index is determined based on the normalized vegetation index and the normalized water index, and the area of ​​pure water body is obtained based on the surface index.

[0009] Based on the pure water body area, determine the time corresponding to the maximum and minimum pure water body area, and determine the time for obtaining the high water level line and the time for obtaining the low water level line.

[0010] When obtaining the low water level line, the proportion of non-pure pixel water bodies is obtained by using the random forest method, microwave data, normalized vegetation index and normalized water body index. The proportion of non-pure pixel water bodies and the water level line of pure water bodies are combined to use the water level line generation model to obtain the low water level line.

[0011] When acquiring high water level lines, a function is established to compare microwave data with the proportion of water bodies. The proportion of non-pure pixel water bodies is obtained using microwave data. The proportion of non-pure pixel water bodies and the water level lines of pure water bodies are combined to use a water level line generation model to obtain high water level lines.

[0012] By combining high and low water levels with a beach identification model, the beach category can be obtained and the beach area can be determined.

[0013] Preferably, before determining the surface index based on the normalized difference between vegetation index and normalized difference between water index, the following steps are also included:

[0014] Remote sensing data of the sandbars to be identified is acquired, and improved normalized difference water index, normalized water index, normalized vegetation index, and backscattering coefficient are determined based on the remote sensing data, specifically as follows:

[0015] The improved normalized differential water index is determined according to the following formula:

[0016]

[0017] Wherein, MNDWI represents the improved normalized differential water index, p(G) represents green band reflectance, and p(MIR) represents mid-infrared band reflectance.

[0018] The normalized water index is determined according to the following formula:

[0019]

[0020] Wherein, NDWI represents the normalized water index, and p(NIR) represents the near-infrared reflectance.

[0021] The normalized vegetation index is determined according to the following formula:

[0022]

[0023] Wherein, NDVI represents the Normalized Difference Vegetation Index, and p(R) represents the red band reflectance;

[0024] The backscattering coefficient is determined according to the following formula:

[0025] σ ij =VV ij ×VH ij ;

[0026] Where, σ ij V represents the backscattering coefficient, VV represents microwave data with vertical transmission and vertical reception polarization, VH represents microwave data with vertical transmission and horizontal reception polarization, i represents the time sequence number, and j represents the grid sequence number.

[0027] Preferably, the surface index is determined based on the normalized difference between vegetation index and normalized difference between water index, and the area of ​​pure water body is obtained based on the surface index, including:

[0028] The surface index is determined according to the following formula:

[0029]

[0030] SI ij δ represents the surface index, and δ represents the empirical threshold for pure water.

[0031] The surface area of ​​the pure water body is determined according to the following formula:

[0032]

[0033] PWS i J represents the area of ​​the pure water body, J represents the number of grids, and b represents the data resolution of the microwave data.

[0034] Preferably, determining the time corresponding to the maximum and minimum pure water body area based on the pure water body area, and determining the high water level acquisition time and low water level acquisition time, includes:

[0035] Obtain the maximum and minimum surface area of ​​pure water.

[0036] Set the time corresponding to the maximum value as the flood season time, and set i = i max Determine i max Time for obtaining high water level;

[0037] Let the time corresponding to the minimum value be the dry season, and let i = i min Determine i min The time for obtaining the low water level line.

[0038] Preferably, when acquiring the low water level line, the proportion of non-pure pixel water bodies is obtained using the random forest method, microwave data, normalized vegetation index, and normalized water body index. The low water level line is then obtained by combining the proportion of non-pure pixel water bodies with the water level line of pure water bodies using a water level line generation model, including:

[0039] Establish a nonlinear functional relationship between input parameters and the proportion of water body:

[0040] PW = f RF (C)+ε;

[0041] Where PW represents the water body percentage, C represents the input vector of the input parameters, and f RF This represents a nonlinear function relating the input parameters to the water body percentage, where the input parameters include MNDWI, NDVI, NDWI, p(G), p(R), and microwave data.

[0042] A random forest model is established, and the proportion of non-pure pixel water bodies is determined based on the random forest model; the expression for the random forest model is:

[0043]

[0044] Where g(PW|C) represents the ensemble decision tree, n represents the number of regression trees, and g i (PW|C) represents a sub-decision tree;

[0045] The time is obtained at the low water level, i = i min At that time, the water level line l0 of the pure water body was obtained by utilizing the distribution of the pure water body;

[0046] Establish a coordinate system with any point on l0 as the origin, the normal as the y-axis, and the direction pointing towards the land as the positive direction, and obtain... Among them, y min This represents the coordinates of a point on the low water level line, where y0 represents i = i min The coordinates of a point on the water level line of a pure water body. denoted by , where 'a' represents the average percentage of water content in the neighboring pixels of a given point, and 'a' represents the data resolution of the random forest model. means i=i min The serial number of the point on the water level line of the pure water body at that time;

[0047] Obtain the low water level data for ten consecutive years, and determine the average low water level as the low water level line for those ten years.

[0048] Preferably, when acquiring the high water level line, a function is established to correlate microwave data with the proportion of water bodies. The proportion of non-pure pixel water bodies is obtained using the microwave data. The high water level line is then acquired by combining the proportion of non-pure pixel water bodies with the pure water level line using a water level line generation model, including:

[0049] Establish a function of backscattering coefficient and water body proportion: PW = p(σ j ), where PW represents the water body percentage, j represents the raster ordinal number, and σ j Indicates the backscattering coefficient;

[0050] The proportion of water in non-pure pixels is determined as a function of the backscattering coefficient and the proportion of water.

[0051] The time is obtained at the high water level, i = i max At that time, the water level line l1 of the pure water body was obtained by utilizing the distribution of the pure water body;

[0052] Establish a coordinate system with any point on l1 as the origin, the normal as the y-axis, and the direction pointing towards the land as the positive direction, and obtain... Among them, y max y1 represents the coordinates of a point on the high water level line, where y1 represents i = i max The coordinates of a point on the water level line of a pure water body. Let b represent the average percentage of water in the pixels adjacent to a point, b represent the data resolution of the microwave data, and τ represent the value of i = i max The serial number of the point on the water level line of the pure water body at that time;

[0053] Obtain the high water level lines for ten consecutive years, and determine the average high water level line as the high water level line.

[0054] Preferably, a sandbar identification model is used in conjunction with high and low water levels to obtain sandbar categories and determine sandbar areas, including:

[0055] When the aforementioned shoal category is a lake island / shoal, the area of ​​the lake island / shoal is determined according to the following formula:

[0056]

[0057] Where S1 represents the area of ​​the island / shoal in the lake, and n represents i = i min The pixel index of the non-pure water body within the water body, where N represents i = i min The total number of non-pure water pixels in the water body, m represents i = i maxThe non-pure water cell index within the water body, M represents i = i max The total number of non-pure water pixels in the water body, and S0 represents the area of ​​permanent water bodies surrounded by islands and shoals in the lake.

[0058] Preferably, a sandbar identification model is used in conjunction with high and low water levels to obtain sandbar categories and determine sandbar areas, including:

[0059] Using time series methods, areas where the land surface type changes between water and land and are not contained within water bodies during a natural year are extracted and identified as strip-shaped sandbars.

[0060] When the shoal type is a strip shoal, the width of the strip shoal is determined based on the low water level and high water level of the strip shoal.

[0061] The area of ​​a strip-shaped sandbar is determined by the following formula:

[0062]

[0063] Where S2 represents the area of ​​the strip-shaped shoal, L max L represents the length of the high water level. min Indicates the length of the low water level line. It indicates the width of the strip-shaped sandbar.

[0064] Preferably, a sandbar identification model is used in conjunction with high and low water levels to obtain sandbar categories and determine sandbar areas, including:

[0065] Obtain the radius of curvature at each point on the low water level line, and establish a function of the water level line and the radius of curvature, R = p(x), where R represents the radius of curvature and x represents the distance from a point on the low water level line to the source of the river;

[0066] Let x be the point where R = 0 and p(x) ≤ r0. g ;

[0067] Establish a point x g With the origin as the starting point, and point x as the starting point... g The tangent to the x-axis is taken as the x-axis, with counterclockwise as the positive direction, and the x-axis is taken as the x-axis. g A Cartesian coordinate system with the y-axis as the normal and the direction pointing towards the shore as the positive direction;

[0068] The boundary of the shoal is fitted with a quadratic curve to determine the shoal category as a convex bank shoal in the water body. Let p(x) g )=r0 and |x dg The smallest point has coordinates (x, y) in a Cartesian coordinate system. d0 y d0 The equation of the quadratic curve is:

[0069]

[0070] The area of ​​the convex bank of the water body is determined by the equation of a quadratic curve as follows:

[0071]

[0072] Where S3 represents the area of ​​the convex bank of the water body, g represents the vertex index, and G represents the maximum value of g.

[0073] Preferably, the sandbar category is obtained by combining the high water level line and the low water level line with the sandbar identification model, and the sandbar area is determined, including: the sandbar area is the sum of the area of ​​the central sandbar, the area of ​​the strip sandbar and the area of ​​the sandbar on the convex bank of the water body, that is, S=S1+S2+S3.

[0074] Compared with existing technologies, the beneficial effects of this invention are as follows: This invention utilizes the distribution and morphological characteristics of sandbars, extracts high and low water levels, and combines remote sensing technology to extract water and land surface features. Through remote sensing, it accurately identifies different types of sandbars and obtains accurate distribution information. This invention comprehensively considers multiple factors such as water vegetation index, surface index of pure water bodies, flood season duration, and dry season duration using remote sensing imagery, making the identification results more comprehensive and reliable. It can promptly acquire information on the sandbars to be identified, enabling real-time monitoring and dynamic analysis of sandbar land area. It is applicable to different types of sandbars, exhibiting strong adaptability and versatility. The remote sensing technology of this invention greatly improves identification efficiency, reduces the tediousness and errors of manual measurement, helps assess the ecological environment status of sandbars, provides a basis for environmental protection, and accurately grasps the land area of ​​sandbars, facilitating better resource management and allocation. Attached Figure Description

[0075] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0076] Figure 1 This is a flowchart illustrating a method for identifying and remotely estimating the area of ​​a beach according to the present invention. Detailed Implementation

[0077] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0078] like Figure 1 As shown, the present invention provides a method for identifying and remotely estimating the area of ​​sandbars, including: determining the surface index based on the normalized vegetation index and the normalized water index, and obtaining the area of ​​pure water bodies based on the surface index.

[0079] Based on the pure water body area, determine the time corresponding to the maximum and minimum pure water body area, and determine the acquisition time of the high water level line and the acquisition time of the low water level line.

[0080] When obtaining the low water level line, the proportion of non-pure pixel water bodies is obtained by using the random forest method, microwave data, normalized vegetation index, and normalized water body index. The water level line generation model is then used to combine the proportion of non-pure pixel water bodies and the water level line of pure water bodies to obtain the low water level line.

[0081] When acquiring high water level lines, a function is established to compare microwave data with the proportion of water bodies. The proportion of water bodies in non-pure pixels is obtained using microwave data. The proportion of water bodies in non-pure pixels and the water level lines in pure water bodies are combined to use a water level line generation model to obtain high water level lines.

[0082] By combining high and low water levels with a beach identification model, the beach category can be obtained and the beach area can be determined.

[0083] In some embodiments of this application, before determining the surface indices based on the normalized difference in vegetation index and the normalized difference in water index, the method further includes: acquiring remote sensing data of the sandbar to be identified, and determining the improved normalized difference in water index, the normalized difference in water index, the normalized difference in vegetation index, and the backscattering coefficient based on the remote sensing data, specifically:

[0084] The improved normalized differential water index is determined according to the following formula:

[0085]

[0086] Wherein, MNDWI represents the improved normalized differential water index, p(G) represents green band reflectance, and p(MIR) represents mid-infrared band reflectance.

[0087] The normalized water index is determined according to the following formula:

[0088]

[0089] Wherein, NDWI represents the normalized water index, and p(NIR) represents the near-infrared reflectance.

[0090] The normalized vegetation index is determined according to the following formula:

[0091]

[0092] Wherein, NDVI represents the Normalized Difference Vegetation Index, and p(R) represents the red band reflectance;

[0093] The backscattering coefficient is determined according to the following formula:

[0094] σ ij =VV ij ×VH ij ;

[0095] Where, σ ij V represents the backscattering coefficient, VV represents microwave data with vertical transmission and vertical reception polarization, VH represents microwave data with vertical transmission and horizontal reception polarization, i represents the time sequence number, and j represents the grid sequence number.

[0096] In this embodiment, microwave data is filtered based on polarization and an ascending imaging orbit. The images are then cropped according to the study area and RefinedLee filtering is used to eliminate some SAR image speckle noise.

[0097] In some embodiments of this application, the surface index is determined based on the normalized vegetation index and the normalized water index, and the area of ​​pure water body is obtained based on the surface index, including:

[0098] The surface index is determined according to the following formula:

[0099]

[0100] SI ij δ represents the surface index, and δ represents the empirical threshold for pure water.

[0101] The surface area of ​​the pure water body is determined according to the following formula:

[0102]

[0103] PWS i J represents the area of ​​the pure water body, J represents the number of grids, and b represents the data resolution of the microwave data.

[0104] In some embodiments of this application, the time corresponding to the maximum and minimum pure water body area is determined based on the pure water body area, and the time for obtaining the high water level line and the low water level line is determined, including: obtaining the maximum and minimum pure water body area; setting the time corresponding to the maximum value as the flood season time, and setting i = i max Determine i max The time for obtaining the high water level is set; the time corresponding to the minimum value is set as the dry season time, and i = i min Determine i min The time for obtaining the low water level line.

[0105] In this embodiment, the time corresponding to the maximum annual area of ​​pure water body is designated as the flood season, and the time corresponding to the minimum annual area of ​​pure water body is designated as the dry season.

[0106] In some embodiments of this application, when obtaining the low water level line, the proportion of non-pure pixel water bodies is obtained using the random forest method, microwave data, normalized vegetation index, and normalized water body index. The low water level line is then obtained by combining the proportion of non-pure pixel water bodies with the water level line of pure water bodies using a water level line generation model, including:

[0107] Establish a nonlinear functional relationship between input parameters and the proportion of water body:

[0108] PW = f RF (C)+ε;

[0109] Where PW represents the water body percentage, C represents the input vector of the input parameters, and f RF This represents a nonlinear function relating the input parameters to the water body percentage, where the input parameters include MNDWI, NDVI, NDWI, p(G), p(R), and microwave data.

[0110] A random forest model is established, and the proportion of non-pure pixel water bodies is determined based on the random forest model; the expression for the random forest model is:

[0111]

[0112] Where g(PW|C) represents the ensemble decision tree, n represents the number of regression trees, and g i (PW|C) represents a sub-decision tree;

[0113] The time is obtained at the low water level, i = i min At that time, the water level line l0 of the pure water body was obtained by utilizing the distribution of the pure water body;

[0114] Establish a coordinate system with any point on l0 as the origin, the normal as the y-axis, and the direction pointing towards the land as the positive direction, and obtain... Among them, y minThis represents the coordinates of a point on the low water level line, where y0 represents i = i min The coordinates of a point on the water level line of a pure water body. denoted by , where 'a' represents the average percentage of water content in the neighboring pixels of a given point, and 'a' represents the data resolution of the random forest model. means i=i min The serial number of the point on the water level line of the pure water body at that time;

[0115] Obtain the low water level data for ten consecutive years, and determine the average low water level as the low water level line for those ten years.

[0116] In this embodiment, existing MNDWI, NDVI, NDWI, p(G), p(R) and microwave data are used as highly correlated influencing factors of water body proportion. The proportion of water body in a grid at a higher resolution is obtained using remote sensing data on the distribution of water and non-water bodies.

[0117] Random forests build multiple decision trees during the training phase and then calculate the average prediction of these decision trees as the output of the method. The input feature space is divided into a large number of regression trees, each generated from a bootstrap sample. A bootstrap sample consists of two-thirds of the training samples, and the remaining one-third of the data is used to validate each tree.

[0118] The influence of random forest ranking parameters on water body proportion is used to rank the data, and the top 4 are selected as key vectors for water body proportion. Inputting these key vectors into the trained model yields a water body proportion map. Then, based on i = i... min When fitting the pure water body distribution to the pure water body water level line l0, a coordinate system is set with any point on l0 as the origin, the normal as the y-axis, and the direction pointing towards the land as the positive direction.

[0119] In some embodiments of this application, when acquiring the high water level line, a function is established between microwave data and water body proportion. The microwave data is used to obtain the proportion of non-pure pixel water bodies. The high water level line is then obtained by combining the proportion of non-pure pixel water bodies and the pure water body water level line using a water level line generation model. This includes:

[0120] Establish a function of backscattering coefficient and water body proportion: PW = p(σ j ), where PW represents the water body percentage, j represents the raster ordinal number, and σ j Indicates the backscattering coefficient;

[0121] The proportion of water in non-pure pixels is determined as a function of the backscattering coefficient and the proportion of water.

[0122] The time is obtained at the high water level, i = i max At that time, the water level line l1 of the pure water body was obtained by utilizing the distribution of the pure water body;

[0123] Establish a coordinate system with any point on l1 as the origin, the normal as the y-axis, and the direction pointing towards the land as the positive direction, and obtain... Among them, y max y1 represents the coordinates of a point on the high water level line, where y1 represents i = i max The coordinates of a point on the water level line of a pure water body. Let b represent the average percentage of water in the pixels adjacent to a point, b represent the data resolution of the microwave data, and τ represent the value of i = i max The serial number of the point on the water level line of the pure water body at that time;

[0124] Obtain the high water level lines for ten consecutive years, and determine the average high water level line as the high water level line.

[0125] In this embodiment, when acquiring the high water level line, the area is in its high-water season with frequent cloud and rain weather, significantly reducing the accuracy and usability of optical remote sensing data. Therefore, microwave remote sensing data is used to acquire the water level line. However, microwave remote sensing data has low resolution, and in the field, the shoreline contains mountains, rocks, vegetation, and other materials, making it impossible to accurately extract the high water level line from the pixels along the shoreline, which are a combination of water and non-water bodies. Therefore, it is necessary to establish a function of backscattering coefficient and water body proportion. Based on i = i max When fitting the pure water body distribution to the pure water body water level line l1, a coordinate system is set with any point on l1 as the origin, the normal as the y-axis, and the direction pointing towards the land as the positive direction.

[0126] In some embodiments of this application, a beach identification model is used in conjunction with high and low water levels to obtain beach categories and determine beach areas, including:

[0127] When the aforementioned shoal category is a lake island / shoal, the area of ​​the lake island / shoal is determined according to the following formula:

[0128]

[0129] Where S1 represents the area of ​​the island / shoal in the lake, and n represents i = i min The pixel index of the non-pure water body within the water body, where N represents i = i min The total number of non-pure water pixels in the water body, m represents i = i max The non-pure water cell index within the water body, M represents i = i max The total number of non-pure water pixels in the water body, and S0 represents the area of ​​permanent water bodies surrounded by islands and shoals in the lake.

[0130] In this embodiment, since there may be a lake island or islet in the center of the lake, the area around the lake island, the permanent water body in the center of the lake island, and the area around the lake island all belong to the shoals / islands. Furthermore, since i = i maxWithin an area surrounded by water, there may be non-pure water pixels; therefore, the sandbar region is defined as i = i min Non-pure water pixel removal i = i max The portion of the non-pure water body pixels and the additional pixels at i=i min The portion of pure water contained within the non-pure water pixels that are included in the water body.

[0131] In some embodiments of this application, a beach identification model is used in conjunction with high and low water levels to obtain beach categories and determine beach areas, including:

[0132] Using time series methods, areas where the land surface type changes between water and land and are not contained within water bodies during a natural year are extracted and identified as strip-shaped sandbars.

[0133] When the shoal type is a strip shoal, the width of the strip shoal is determined based on the low water level and high water level of the strip shoal.

[0134] The area of ​​a strip-shaped sandbar is determined by the following formula:

[0135]

[0136] Where S2 represents the area of ​​the strip-shaped shoal, L max L represents the length of the high water level. min Indicates the length of the low water level line. It indicates the width of the strip-shaped sandbar.

[0137] In this embodiment, the strip-shaped shoal is the most common type of shoal, typically distributed around water bodies and formed by annual hydrological rhythms and water level changes. When extracting the strip-shaped shoal, the area of ​​the strip-shaped shoal region is obtained using the length of the water level line and the width of the shoal. Using time series methods, areas where the surface type changes between water and land within a natural year and are not contained within the water body are extracted; these areas are the strip-shaped shoals. A normal line is drawn from a point on the low water level line, intersecting the high water level line at a single point. The distance between these two points is the width of the strip-shaped shoal. The width of the strip-shaped shoal is taken at intervals of K, and the average value is calculated.

[0138] In some embodiments of this application, a beach identification model is used in conjunction with high and low water levels to obtain beach categories and determine beach areas, including:

[0139] Obtain the radius of curvature at each point on the low water level line, and establish a function of the water level line and the radius of curvature, R = p(x), where R represents the radius of curvature and x represents the distance from a point on the low water level line to the source of the river;

[0140] Let x be the point where R = 0 and p(x) ≤ r0. g ;

[0141] Establish a point x g With the origin as the starting point, and point x as the starting point... g The tangent to the x-axis is taken as the x-axis, with counterclockwise as the positive direction, and the x-axis is taken as the x-axis. g A Cartesian coordinate system with the y-axis as the normal and the direction pointing towards the shore as the positive direction;

[0142] The boundary of the shoal is fitted with a quadratic curve to determine the shoal category as a convex bank shoal in the water body. Let p(x) g )=r0 and |x dg The smallest point has coordinates (x, y) in a Cartesian coordinate system. d0 y d0 The equation of the quadratic curve is:

[0143]

[0144] The area of ​​the convex bank of the water body is determined by the equation of a quadratic curve as follows:

[0145]

[0146] Where S3 represents the area of ​​the convex bank of the water body, g represents the vertex index, and G represents the maximum value of g.

[0147] In this embodiment, when setting the origin, the point where R = 0 and p(x) ≤ r0 (r0 is the threshold) is taken as x. g Let g be the vertex index, and let G be the maximum value of g. Establish a vertex index x... g A Cartesian coordinate system with the origin at point X, the tangent at point X as the x-axis, counterclockwise as the positive direction, the normal at point X as the y-axis, and the direction pointing towards the shore as the positive direction.

[0148] In some embodiments of this application, the high water level line and low water level line are combined with a beach identification model to obtain the beach category and determine the beach area, including: the beach area is the sum of the area of ​​the lake center beach, the area of ​​the strip beach and the area of ​​the beach on the convex bank of the water body, that is, S = S1 + S2 + S3.

[0149] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

[0150] The system provided in the above embodiments is only illustrated by the division of the above functional modules. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the modules or steps in the embodiments of the present invention can be further decomposed or combined. For example, the modules in the above embodiments can be merged into one module, or further divided into multiple sub-modules to complete all or part of the functions described above. The names of the modules and steps involved in the embodiments of the present invention are only for distinguishing the various modules or steps and are not considered as an improper limitation of the present invention.

[0151] Those skilled in the art will recognize that the modules and method steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. The programs corresponding to the software modules and method steps can be placed in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disks, removable disks, CD-ROMs, or any other form of storage medium known in the art. To clearly illustrate the interchangeability of electronic hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in electronic hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the invention.

Claims

1. A method for identifying and remotely estimating the area of ​​sandbars, characterized in that, include: The surface index is determined based on the normalized vegetation index and the normalized water index, and the area of ​​pure water body is obtained based on the surface index. Based on the pure water body area, determine the time corresponding to the maximum and minimum pure water body area, and determine the time for obtaining the high water level line and the time for obtaining the low water level line. When obtaining the low water level line, the proportion of non-pure pixel water bodies is obtained by using the random forest method, microwave data, normalized vegetation index and normalized water body index. The proportion of non-pure pixel water bodies and the water level line of pure water bodies are combined to use the water level line generation model to obtain the low water level line. When acquiring high water level lines, a function is established to compare microwave data with the proportion of water bodies. The proportion of non-pure pixel water bodies is obtained using microwave data. The proportion of non-pure pixel water bodies and the water level lines of pure water bodies are combined to use a water level line generation model to obtain high water level lines. By combining high and low water levels with a beach identification model, the beach category is obtained, and the beach area is determined, including: When the aforementioned shoal category is a lake island / shoal, the area of ​​the lake island / shoal is determined according to the following formula: Where S1 represents the area of ​​the island / shoal in the lake, and n represents i = The pixel number of the non-pure water body within the water body, where N represents i= The total number of non-pure water pixels in the water body, m represents i= The pixel number of the non-pure water body within the water body, M represents i= The total number of non-pure water pixels in the water body, S0 represents the area of ​​permanent water bodies surrounded by islands and shoals in the lake center; To obtain the time for the high water level line, For obtaining the low water level; Using time series analysis, areas where land surface types change between water and land bodies but are not contained within water bodies during a natural year are extracted and identified as strip-shaped sandbars. When a sandbar is classified as a strip-shaped sandbar, its width is determined based on its low and high water levels. ; The area of ​​a strip-shaped sandbar is determined by the following formula: Where S2 represents the area of ​​the strip-shaped sandbar, Indicates the length of the high water level line. Indicates the length of the low water level line. Indicates the width of a strip-shaped sandbar; Obtain the radius of curvature at each point on the low water level line, and establish a function R = p(x) relating the water level line and the radius of curvature, where R represents the radius of curvature and x represents the distance from a point on the low water level line to the river source; define the point corresponding to R = 0 and p(x) ≤ r0 as... Establish a point-based system With the origin as the starting point, and the point as the starting point The tangent to the x-axis is taken as the x-axis, with counterclockwise as the positive direction, and the point is... A Cartesian coordinate system with the normal to the y-axis and the direction pointing towards the shore as the positive direction; The boundary of the shoal is fitted with a quadratic curve to determine the shoal category as a convex bank shoal in the water body. Let p( )=r0 and The coordinates of the smallest point in the Cartesian coordinate system are: The equation of the quadratic curve is: The area of ​​the convex bank of the water body is determined by the equation of a quadratic curve as follows: in, Let g represent the area of ​​the convex bank of the water body, g represent the vertex index, and G represent the maximum value of g.

2. The method for identifying and remotely estimating the area of ​​sandbars according to claim 1, characterized in that, Before determining the surface index based on the normalized difference between vegetation index and normalized difference between water index, the following steps are also included: Remote sensing data of the sandbars to be identified is acquired, and improved normalized difference water index, normalized water index, normalized vegetation index, and backscattering coefficient are determined based on the remote sensing data, specifically as follows: The improved normalized differential water index is determined according to the following formula: Wherein, MNDWI represents the improved normalized differential water index, p(G) represents green band reflectance, and p(MIR) represents mid-infrared band reflectance. The normalized water index is determined according to the following formula: Wherein, NDWI represents the normalized water index, and p(NIR) represents the near-infrared reflectance. The normalized vegetation index is determined according to the following formula: Wherein, NDVI represents the Normalized Difference Vegetation Index, and p(R) represents the red band reflectance; The backscattering coefficient is determined according to the following formula: in, V represents the backscattering coefficient, VV represents microwave data with vertical transmission and vertical reception polarization, VH represents microwave data with vertical transmission and horizontal reception polarization, i represents the time sequence number, and j represents the grid sequence number.

3. The method for identifying and remotely estimating the area of ​​sandbars according to claim 2, characterized in that, Surface indices are determined based on the normalized difference between vegetation index and normalized difference between water index, and the area of ​​pure water bodies is obtained based on the surface indices, including: The surface index is determined according to the following formula: Indicates surface index, An empirical threshold representing a pure water volume; The surface area of ​​the pure water body is determined according to the following formula: J represents the area of ​​the pure water body, J represents the number of grids, and b represents the data resolution of the microwave data.

4. The method for identifying and remotely estimating the area of ​​sandbars according to claim 3, characterized in that, Based on the pure water volume area, determine the time corresponding to the maximum and minimum pure water volume areas, and determine the acquisition time for the high water level line and the low water level line, including: Obtain the maximum and minimum surface area of ​​pure water. Set the time corresponding to the maximum value as the flood season time, and set i= ,Sure Time for obtaining high water level; Let the time corresponding to the minimum value be the dry season, and let i = ,Sure The time for obtaining the low water level line.

5. The method for identifying and remotely estimating the area of ​​sandbars according to claim 4, characterized in that, When acquiring low water level lines, the random forest method, microwave data, normalized vegetation index, and normalized water index are used to obtain the proportion of non-pure pixel water bodies. The proportion of non-pure pixel water bodies and the water level lines of pure water bodies are then combined to use a water level line generation model to obtain the low water level lines, including: Establish a nonlinear functional relationship between input parameters and the proportion of water body: Where PW represents the water body percentage, and C represents the input vector of the input parameters. This represents a nonlinear function relating the input parameters to the water body percentage, where the input parameters include MNDWI, NDVI, NDWI, p(G), p(R), and microwave data. A random forest model is established, and the proportion of non-pure pixel water bodies is determined based on the random forest model; the expression for the random forest model is: Where g(PW|C) represents the ensemble decision tree, and n represents the number of regression trees. Representing a sub-decision tree; The time is obtained at the low water level, i.e., i= At that time, the water level of the pure water body was obtained by utilizing the distribution of the pure water body. ; by Establish a coordinate system with any point as the origin, the normal as the y-axis, and the direction pointing towards the land as the positive direction, and obtain... ;in, This represents the coordinates of a point on the low water level line. Indicate i= The coordinates of a point on the water level line of a pure water body. denoted by , where 'a' represents the average percentage of water content in the neighboring pixels of a given point, and 'a' represents the data resolution of the random forest model. Indicate i= The serial number of the point on the water level line of the pure water body at that time; Obtain the low water level data for ten consecutive years, and determine the average low water level as the low water level line for those ten years. .

6. The method for identifying and remotely estimating the area of ​​sandbars according to claim 5, characterized in that, When acquiring high water level lines, a function is established to correlate microwave data with water body proportions. Microwave data is used to obtain the proportion of non-pure pixel water bodies. The high water level lines are then generated using a water level line generation model, combining the proportion of non-pure pixel water bodies with the water level lines of pure water bodies. This includes: Establish a function relating the backscattering coefficient to the water body percentage: Where PW represents the water body percentage and j represents the raster ordinal number. Indicates the backscattering coefficient; The proportion of water in non-pure pixels is determined as a function of the backscattering coefficient and the proportion of water. The time is obtained at the high water level, i.e., i= At that time, the water level of the pure water body was obtained by utilizing the distribution of the pure water body. ; by Establish a coordinate system with any point as the origin, the normal as the y-axis, and the direction pointing towards the land as the positive direction, and obtain... in, This represents the coordinates of a point on the high water level line. Indicate i= The coordinates of a point on the water level line of a pure water body. 'b' represents the average percentage of water content in the pixels adjacent to a given point, and 'b' represents the data resolution of the microwave data. Indicate i= The serial number of the point on the water level line of the pure water body at that time; Obtain the high water level lines for ten consecutive years, and determine the average high water level line as the high water level line. .

7. The method for identifying and remotely estimating the area of ​​sandbars according to claim 6, characterized in that, By combining high and low water levels with a beach identification model, the beach category is obtained and the beach area is determined, including: the beach area is the sum of the area of ​​the lake center beach, the area of ​​the strip beach, and the area of ​​the beach on the convex bank of the water body, i.e., S=S1+S2+S3.

Images