A soil wind erosion monitoring method and system based on remote sensing image data

Abstract

The present application relates to the technical field of image processing, and particularly relates to a soil wind erosion monitoring method and system based on remote sensing image data. The method comprises the following steps: collecting remote sensing image data, regional wind field data and terrain elevation data corresponding to a target monitoring area; determining the consistency of ground texture direction according to the remote sensing image data; filtering the particle migration direction based on the consistency of ground texture direction and the regional wind field data; establishing the regional wind force action relationship according to the regional wind field data and the terrain elevation data, and identifying the wind erosion activity area in combination with the particle migration direction; detecting the effective wind erosion area based on the ground particle migration track and the vegetation coverage change in each wind erosion activity area, and dividing the wind erosion grade; generating the soil wind erosion spatial distribution result corresponding to the target monitoring area, and constructing the regional wind erosion evolution sequence. The present application can improve the soil wind erosion monitoring efficiency and improve the capture ability of dynamic changes of erosion.

Description

Technical Field

[0001] This invention relates to the field of image processing technology, and in particular to a method and system for monitoring soil wind erosion based on remote sensing image data. Background Technology

[0002] Soil wind erosion is a common form of land degradation in arid, semi-arid, and ecologically fragile regions. It is primarily characterized by the stripping, transport, and deposition of surface soil particles under wind action, leading to a thinning of the topsoil, decreased soil fertility, altered surface roughness, and weakened regional ecosystem stability. Particularly in sandy areas, desert fringe zones, agro-pastoral ecotones, and seasonally exposed farmland, soil wind erosion processes exhibit significant dynamics and spatial heterogeneity due to changes in vegetation cover, fluctuations in surface moisture, and human farming activities.

[0003] Existing technologies for monitoring soil wind erosion mainly include manual field surveys, sampling at fixed observation stations, and empirical model analysis based on meteorological parameters. Manual field surveys typically assess the degree of wind erosion by collecting soil particles, measuring surface wind erosion grooves, and analyzing changes in land cover. However, this method suffers from limited monitoring range, long cycles, high manpower requirements, and difficulty in achieving continuous dynamic monitoring. While fixed-station monitoring can obtain information on wind speed, sediment transport, and changes in surface particles in local areas, the limited number of monitoring points makes it difficult to accurately reflect the spatial distribution differences of wind erosion over large-scale regions. Empirical models built based on meteorological data can achieve regional-scale analysis, but they usually rely on preset empirical parameters and have weak responsiveness to complex factors such as land cover status, soil moisture content, and topographic changes, leading to significant discrepancies between monitoring results and actual erosion conditions. Summary of the Invention

[0004] Therefore, it is necessary for the present invention to provide a method and system for monitoring soil wind erosion based on remote sensing image data, in order to solve at least one of the above-mentioned technical problems.

[0005] To achieve the above objectives, a method for monitoring soil wind erosion based on remote sensing image data includes the following steps:

[0006] Step S1: Collect remote sensing image data, regional wind field data, and topographic elevation data corresponding to the target monitoring area;

[0007] Step S2: Determine the consistency of surface texture direction based on the differences in texture gradient change characteristics between corresponding feature regions of adjacent image frames in the remote sensing image data; identify particle migration traces based on the consistency of surface texture direction and regional wind field data to filter particle migration directions.

[0008] Step S3: Establish the regional wind force relationship based on regional wind field data and topographic elevation data, and detect the migration trajectory of surface particles in combination with particle migration direction to identify areas of wind erosion activity;

[0009] Step S4: Based on the migration trajectory of surface particles and changes in vegetation cover in each wind erosion activity area, detect the effective wind erosion area and classify the wind erosion level corresponding to each effective wind erosion area.

[0010] Step S5: Generate the spatial distribution results of soil wind erosion in the target monitoring area based on the wind erosion level, and construct the regional wind erosion evolution sequence by combining the changes in wind erosion level at different time points.

[0011] This application utilizes a multi-source collaborative analysis mechanism combining remote sensing imagery data, regional wind field data, and topographic elevation data to achieve continuous, dynamic, and large-scale monitoring of soil wind erosion processes. Compared to traditional manual survey methods, this significantly reduces reliance on field sampling and manual interpretation, thereby lowering monitoring costs and improving efficiency. Simultaneously, by analyzing texture gradient changes and directional texture bands based on remote sensing images at adjacent time points, it enables automatic identification of surface particle migration traces, transforming wind erosion process assessment from static evaluation to temporal evolution analysis, thus enhancing the ability to capture dynamic changes in the erosion process. Furthermore, it integrates wind field propagation paths and topographic relief... The feature-based regional wind action relationship is constructed to establish a constraint mapping relationship between particle migration direction and wind-driven mechanism, thereby improving the physical consistency and spatial interpretability of wind erosion identification. In addition, through the spatial intersection screening mechanism of vegetation cover change and wind erosion activity area, the interference of surface change caused by non-wind factors can be effectively eliminated, improving the accuracy and robustness of effective wind erosion area identification. Finally, by constructing wind erosion intensity classification and temporal evolution sequence, the spatial distribution and development trend of regional wind erosion can be quantitatively expressed, providing higher precision and stronger spatiotemporal continuity technical support for ecological degradation assessment and land governance decision-making.

[0012] Optionally, this application also provides a soil wind erosion monitoring system based on remote sensing image data, used to execute the soil wind erosion monitoring method based on remote sensing image data as described above. The soil wind erosion monitoring system based on remote sensing image data includes:

[0013] The data acquisition module is used to collect remote sensing image data, regional wind field data, and terrain elevation data corresponding to the target monitoring area;

[0014] The wind erosion feature determination module is used to determine the consistency of surface texture direction based on the differences in texture gradient change features between corresponding feature regions of adjacent image frames in remote sensing image data; and to identify particle migration traces based on the consistency of surface texture direction and regional wind field data in order to filter particle migration direction.

[0015] The activity area identification module is used to establish the relationship between regional wind forces based on regional wind field data and topographic elevation data, and to detect the migration trajectory of surface particles by combining the particle migration direction corresponding to different time nodes, so as to identify the wind erosion activity area.

[0016] The erosion level classification module is used to detect effective wind erosion areas based on the migration trajectory of surface particles and changes in vegetation cover in each wind erosion activity area, and to classify the wind erosion level corresponding to each effective wind erosion area.

[0017] The erosion evolution analysis module is used to generate the spatial distribution results of soil wind erosion in the target monitoring area based on the wind erosion level, and to construct the regional wind erosion evolution sequence by combining the changes in wind erosion level at different time points.

[0018] The soil wind erosion monitoring system based on remote sensing image data of the present invention can implement any of the soil wind erosion monitoring methods based on remote sensing image data of the present invention. It serves as a medium for coordinating the operation and signal transmission between various modules to complete the soil wind erosion monitoring method based on remote sensing image data. The modules within the system cooperate with each other to achieve a quantitative expression of the spatial distribution and development trend of regional wind erosion, providing higher accuracy and stronger spatiotemporal continuity of technical support for ecological degradation assessment and land management decision-making. Attached Figure Description

[0019] Other features, objects, and advantages of the invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0020] Figure 1 This is a schematic diagram of the steps of the soil wind erosion monitoring method based on remote sensing image data of the present invention;

[0021] Figure 2 This is a schematic diagram of the characteristic region division results of the target monitoring area in one embodiment;

[0022] Figure 3 This is a map showing the distribution of wind erosion levels in the target monitoring area in one embodiment;

[0023] Figure 4 This is a block diagram of the soil wind erosion monitoring system based on remote sensing image data of the present invention;

[0024] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0025] The technical method of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0026] Furthermore, the accompanying drawings are merely illustrative of the invention and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and therefore repeated descriptions of them will be omitted. Some block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities. These functional entities can be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor methods and / or microcontroller methods.

[0027] It should be understood that although the terms "first," "second," etc., may be used herein to describe various units, these units should not be limited by these terms. These terms are used merely to distinguish one unit from another. For example, without departing from the scope of the exemplary embodiments, a first unit may be referred to as a second unit, and similarly, a second unit may be referred to as a first unit. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0028] To achieve the above objectives, please refer to Figures 1 to 4 This invention provides a method for monitoring soil wind erosion based on remote sensing image data, the method comprising the following steps:

[0029] Step S1: Collect remote sensing image data, regional wind field data, and topographic elevation data corresponding to the target monitoring area;

[0030] In one embodiment of this application, multi-temporal remote sensing image data of the target monitoring area acquired over 15 consecutive days are first retrieved and uniformly converted into 10m spatial resolution raster data. Regional wind field data uses wind speed, direction, and frequency records within the corresponding time period, with a time interval of 1 hour. Topographic elevation data uses a 30m precision digital elevation raster, and slope and aspect parameters are further calculated. Subsequently, radiometric correction, cloud occlusion removal, and time registration are performed on the remote sensing images to control the pixel offset error between adjacent image frames to within one pixel, thereby generating basic analysis data of the target monitoring area at a unified time scale.

[0031] Step S2: Determine the consistency of surface texture direction based on the differences in texture gradient change characteristics between corresponding feature regions of adjacent image frames in the remote sensing image data; identify particle migration traces based on the consistency of surface texture direction and regional wind field data to filter particle migration directions.

[0032] In a further embodiment, texture gradient analysis is first performed on remote sensing images at consecutive time points using a 9×9 pixel neighborhood window. The proportion of grayscale change directions of each pixel within the window is statistically analyzed, and directions with a proportion exceeding 60% are identified as the main gradient directions. If the offset angle of the main gradient direction between adjacent time points is less than 15° and the gradient amplitude continuously increases, it is determined that the corresponding feature region has consistent surface texture direction. Subsequently, edge tracking and displacement detection are performed on texture bands with consistent directions. The spatial displacement distance of each texture band within consecutive time points is statistically analyzed, and texture band directions with an angle less than 20° with the prevailing wind direction are selected as candidate particle migration directions. Finally, direction response allocation processing is performed based on the texture band coverage, texture band density, and wind erosion direction response parameters to determine the particle migration direction at the corresponding time point.

[0033] Step S3: Establish the regional wind force relationship based on regional wind field data and topographic elevation data, and detect the migration trajectory of surface particles in combination with particle migration direction to identify areas of wind erosion activity;

[0034] In a further embodiment, wind speed and wind frequency variation parameters are first extracted based on the dominant wind direction corresponding to each characteristic region. The wind speed is the average wind speed over a continuous 6-hour period, and the wind frequency variation parameter is the frequency of the prevailing wind direction within 24 hours. Subsequently, the continuity of wind propagation is analyzed by combining slope, aspect, and topographic relief parameters, and the regional wind transport intensity parameters are corrected based on the topographic shielding relationship between the windward and leeward regions. For regions with a slope exceeding 25° and topographic relief greater than 12m, the wind propagation weight of the corresponding region is reduced. Then, the deflection difference between the particle migration direction and the wind propagation path direction is statistically analyzed. If the deflection difference is less than 20° and the particle migration expansion continuously increases, it is determined that the corresponding region has a continuous surface particle migration trajectory and is further identified as a wind erosion activity area.

[0035] Step S4: Based on the migration trajectory of surface particles and changes in vegetation cover in each wind erosion activity area, detect the effective wind erosion area and classify the wind erosion level corresponding to each effective wind erosion area.

[0036] In a further embodiment, the vegetation cover rate of each characteristic region is first calculated using the reflectance of the red and near-infrared bands in remote sensing images, and the rate of decline in vegetation cover rate within consecutive time nodes is statistically analyzed. When the vegetation cover rate declines by more than 8% continuously, the corresponding region is classified as a vegetation degradation region. Subsequently, spatial intersection analysis is performed on the wind erosion activity region and the vegetation degradation region. When the spatial overlap ratio exceeds 70%, the corresponding region is identified as a valid wind erosion region. Then, a weighted calculation is performed based on the surface particle migration trajectory, the rate of decline in vegetation cover rate, and the amplitude of regional brightness change, where the particle migration trend weight is set to 0.5, the vegetation cover change weight is set to 0.3, and the brightness change weight is set to 0.2. The wind erosion levels are then classified as mild, moderate, severe, and extremely severe according to the erosion level correspondence table.

[0037] Step S5: Generate the spatial distribution results of soil wind erosion in the target monitoring area based on the wind erosion level, and construct the regional wind erosion evolution sequence by combining the changes in wind erosion level at different time points.

[0038] In a further embodiment, a regional wind erosion level distribution map is first generated according to the wind erosion level corresponding to each effective wind erosion area, and the spatial distribution result of soil wind erosion in the target monitoring area is constructed based on the area proportion corresponding to different levels. Subsequently, the changes in wind erosion level of each area within consecutive time nodes are statistically analyzed over time. When the erosion level between adjacent time nodes increases by more than one level, the wind erosion enhancement time node for the corresponding area is recorded. Then, the spatial distribution results of wind erosion level corresponding to each time node are superimposed in chronological order to generate a regional wind erosion evolution sequence corresponding to the target monitoring area, which is used to reflect the expansion and attenuation process of wind erosion activity in different time stages.

[0039] Figure 3 This is a map showing the distribution of wind erosion levels in the target monitoring area in one embodiment; such as... Figure 3 As shown in the figure, light yellow represents mild wind erosion, orange-yellow represents moderate wind erosion, brown-red represents severe wind erosion, and dark brown represents extremely severe wind erosion.

[0040] Optionally, the methods for dividing the feature regions include:

[0041] Extract the surface remote sensing response features corresponding to the target monitoring area from the remote sensing image data, and divide the surface remote sensing response features into several initial surface areas based on the differences in spectral variation and spatial texture distribution.

[0042] In this embodiment, band normalization is first performed on the remote sensing image corresponding to the target monitoring area, and surface remote sensing response features, including surface brightness, texture gradient, and local gray-level variation characteristics, are extracted from the normalized multispectral image. The texture gradient is calculated using a 7×7 pixel sliding window, and brightness variation is represented by the difference in the mean gray-level of adjacent pixels. Subsequently, region segmentation is performed based on the spectral differences and texture distribution differences between regions. When the spectral difference between adjacent regions is less than 0.12 and the texture gradient change rate is less than 15%, the corresponding pixels are assigned to the same initial surface region.

[0043] Based on the regional wind field data, the wind field direction of each initial surface area is detected to determine the distribution of wind field direction in each initial surface area; based on the distribution of wind field direction, the main direction of regional wind field in each initial surface area is clustered and statistically analyzed.

[0044] In a further embodiment, wind field data for each initial surface region at corresponding time points are first retrieved, and the frequency of each wind direction over a continuous 6-hour period is statistically analyzed; the wind direction with the highest frequency is taken as the wind field direction. Then, directional clustering statistics are performed on the wind field directions (i.e., the distribution of wind field directions) within consecutive time points. When the angle between multiple wind directions is less than 20°, they are grouped into the same wind direction category, and the wind direction category with the highest frequency is determined as the dominant wind direction in the region, reflecting the dominant wind direction within the corresponding initial surface region.

[0045] Based on the topographic elevation data, the slope aspect parameters corresponding to each initial surface area are extracted, and the continuous wind propagation status corresponding to each initial surface area is detected based on the spatial correspondence between the slope aspect parameters and the main direction of the regional wind field, thereby identifying the continuous wind propagation path.

[0046] In a further embodiment, the slope aspect parameters of each initial surface area are first calculated based on the digital elevation grid in the terrain elevation data, and the spatial angle between the slope aspect parameters and the main direction of the regional wind field is analyzed. When the angle between the slope aspect and the main direction of the regional wind field is less than 30°, it is determined that the corresponding area has a windward propagation trend. Subsequently, the continuous state of wind propagation is analyzed by combining the elevation difference and slope aspect changes between adjacent areas. Specifically, when the elevation difference between adjacent areas is less than 8m and the slope aspect offset angle is less than 15°, and the directional deflection difference between the main directions of the regional wind field is less than 20°, it is determined that there is a continuous propagation trend of wind in the corresponding area, i.e., a continuous wind propagation state. Subsequently, a regional wind propagation connection relationship is established for adjacent areas that satisfy the continuous propagation trend, and the corresponding connection relationship is used as the continuous wind propagation path.

[0047] If the difference in surface remote sensing response characteristics between adjacent initial surface areas is less than a preset difference threshold, and there is a continuous wind propagation path, then the corresponding adjacent initial surface areas will be merged, and the merged area will be used as the feature area.

[0048] In a further embodiment, the differences in surface remote sensing response characteristics between adjacent initial surface areas are first statistically analyzed. These differences include regional average brightness differences, texture gradient differences, and local grayscale variation differences. If all differences are less than a preset difference threshold of 0.18, and there are continuous wind propagation paths between the corresponding areas, then the corresponding areas undergo regional boundary fusion processing. Afterward, the regional average spectral characteristics and regional texture characteristics are recalculated for the fused areas, and the updated regional results are used as the feature regions corresponding to subsequent wind erosion analysis.

[0049] Figure 2 This is a schematic diagram of the characteristic region division results of the target monitoring area in one embodiment; as shown. Figure 2 As shown in the figure, the areas enclosed by red lines are the feature areas.

[0050] Optionally, determining the consistency of surface texture direction in step S2 includes:

[0051] Based on the differences in texture gradient change characteristics, determine the direction and intensity of grayscale gradient changes between pixels within a preset neighborhood window.

[0052] The main gradient direction of the corresponding neighborhood window is determined based on the proportion of gray-level gradient change directions, and the gradient magnitude of the corresponding neighborhood window is calculated based on the intensity of each gradient change.

[0053] In this embodiment, grayscale gradient calculation is first performed on remote sensing images at adjacent time points using a 9×9 pixel neighborhood window, and the grayscale change direction of each pixel within the window is statistically analyzed in the horizontal, vertical, and diagonal directions. The grayscale gradient change direction is calculated using the grayscale difference between adjacent pixels, and the gradient change intensity is represented by the absolute value of the corresponding grayscale difference. Then, the number of pixels in each grayscale gradient change direction within the same neighborhood window is statistically analyzed. When the number of pixels corresponding to a certain direction accounts for more than 55% of the total number of pixels in the window, that direction is determined as the main gradient direction of the corresponding neighborhood window, and the gradient magnitude of the corresponding neighborhood window is calculated based on the average of all gradient change intensities within the neighborhood window.

[0054] The consistency of texture change direction is detected based on the directional offset between the main gradient directions corresponding to adjacent time nodes, and the texture structure change state is determined by combining the trend of gradient magnitude change.

[0055] Feature regions whose directional offset is less than a preset directional offset threshold and whose texture structure changes meet the preset directional continuous expansion condition are identified as having consistent surface texture direction.

[0056] In a further embodiment, directional offset analysis is performed on the principal gradient direction of the corresponding neighborhood window within consecutive time nodes, and the offset angle of the principal gradient direction between each time node is calculated. This offset angle is the directional offset amount. When the directional offset angle between two consecutive time nodes is less than 15°, it is determined that the corresponding region has a continuous trend of texture direction change. Subsequently, the texture structure change state is analyzed by combining the changes in the gradient amplitude corresponding to each time node of the corresponding neighborhood window. Specifically, when the gradient amplitude change rate within consecutive time nodes remains within 10%, and the gradient amplitude continuously expands along the principal gradient direction, it is determined that the corresponding region has a continuous texture structure change state. Then, feature regions with a principal gradient direction offset angle of less than 15° and whose corresponding texture structure change state meets the preset condition of continuous directional expansion are identified as having consistent surface texture direction.

[0057] It is worth noting that the continuous expansion of gradient magnitude along the principal gradient direction means that, within consecutive time nodes, high gradient regions in the neighborhood window continuously extend in the same direction along the principal gradient direction. For example, in the first time node, high gradient regions are mainly distributed in the middle of the window. In the second time node, if the high gradient regions expand southeastward by 2 to 4 pixels along the principal gradient direction, and the expanded gradient magnitude still maintains more than 85% of the original gradient magnitude, it indicates that the corresponding texture structure has a continuous expansion trend in this direction, reflecting the continuous migration of surface particles along a fixed direction under the action of wind.

[0058] It is worth noting that the preset directional continuous expansion condition means that the principal gradient direction remains basically consistent within consecutive time nodes, and the expansion distance of the high gradient region in the principal gradient direction meets the requirement of continuous growth. For example, if the principal gradient direction offset angle between three consecutive time nodes is less than 15°, and the cumulative expansion distance of the high gradient region in the principal gradient direction exceeds 5 pixels, and the expansion area growth rate exceeds 12%, then the corresponding texture structure is determined to meet the preset directional continuous expansion condition.

[0059] Optionally, identifying particle migration traces in step S2 includes:

[0060] Extract directional consistent texture bands from image frames based on the surface texture direction consistency determination results;

[0061] In this embodiment, the remote sensing image is first processed by directional texture extraction based on the surface texture direction consistency determination result. Specifically, texture regions with a principal gradient direction offset angle of less than 15° and a continuous texture length of more than 12 pixels are extracted as directional consistent texture bands. Then, edge contour tracking is performed on each directional consistent texture band, and the center position coordinates, extension direction and strip length of the corresponding texture band are recorded to generate the spatial distribution result of the texture band at the corresponding time node.

[0062] Perform position offset detection on the corresponding texture bands with consistent orientation in adjacent image frames, and calculate the spatial displacement direction and displacement distance of each texture band with consistent orientation.

[0063] If the spatial displacement direction of the texture bands with consistent orientation arranged continuously along similar directions in adjacent image frames satisfies the preset orientation concentration condition, and the displacement distance satisfies the preset increasing change condition within consecutive time nodes, then it is determined that the texture band with consistent orientation has traces of particle migration.

[0064] In a further embodiment, positional offset detection is performed on texture bands with consistent orientations in adjacent image frames. Using the center coordinates of the texture band in the previous image frame as a reference, a search region with a radius of 8 pixels is established in subsequent image frames. If a corresponding texture band with an extension direction deviation of less than 20° exists within the search region, the spatial displacement direction and displacement distance between the two image frames are calculated. Then, directional aggregation statistics are performed on texture bands with consistent orientations arranged continuously along similar directions. When the spatial displacement directions of multiple texture bands are concentrated within ±15° of the same direction, and the displacement distance gradually increases within three consecutive image frames, it is determined that all texture bands with consistent orientations exhibit particle migration traces.

[0065] Optionally, after determining that there are traces of particle migration in the corresponding feature region, the process also includes:

[0066] Based on the spectral characteristics of each feature region, determine the brightness variation amplitude of the corresponding texture band with consistent direction in each feature region;

[0067] Based on the changes in the edge position and coverage of the directional texture bands within consecutive time nodes, the expansion range of the directional texture bands with particle migration traces is determined, and the activity level of the particle migration direction is evaluated based on the expansion range change status.

[0068] In this embodiment, the average grayscale value change of the consistent texture band corresponding to each feature region is first statistically analyzed within consecutive time nodes, and the brightness change amplitude of the region is calculated based on the grayscale difference between adjacent time nodes. When the average brightness change amplitude of the region corresponding to the consistent texture band exceeds 12%, it is determined that the corresponding region has a significant change in surface reflection. Then, the edge contour of the consistent texture band is continuously tracked, and the changes in the band edge expansion distance and band coverage area within consecutive time nodes are statistically analyzed. If the band coverage area growth rate is continuously greater than 10% within three consecutive time nodes, it is determined that the corresponding consistent texture band has a continuously expanding range, and the activity level of particle migration direction is evaluated based on the band expansion speed in each direction. The band expansion speed is used to assess this activity. It can be represented as: ;in This indicates the area covered by the strip at the current time point. This indicates the area covered by the strip at the previous time point. This represents the time interval between adjacent time nodes. Then, the strip expansion speed is compared with a preset activity level interval to assess the activity level of the particle migration direction. Specifically, when the strip expansion speed is less than 0.5 pixels / h, the corresponding direction is classified as low-activity particle migration and assigned a value of 0.2; when the strip expansion speed is between 0.5 and 1.5 pixels / h, the corresponding direction is classified as medium-activity particle migration and assigned a value of 0.3; and when the strip expansion speed is greater than 1.5 pixels / h, the corresponding direction is classified as high-activity particle migration and assigned a value of 0.5.

[0069] The main direction of the regional wind field at the corresponding time node of the image frame is extracted from the regional wind field data. The wind erosion direction response parameters are calculated based on the degree of direction matching between the extension direction of the consistent texture band and the main wind direction of the regional wind field.

[0070] The distribution of texture bands with consistent direction of particle migration traces was statistically analyzed to determine the density of texture bands, and the surface wind erosion intensity variation parameters were calculated by combining the amplitude of regional brightness variation and the activity of particle migration direction.

[0071] The particle migration direction activity level, wind erosion direction response parameters, and surface wind erosion intensity change parameters are weighted and evaluated to screen the particle migration direction corresponding to the characteristic areas at each time point.

[0072] In a further embodiment, the main direction of the regional wind field corresponding to each time node is extracted from the regional wind field data, and the directional angle between each extension direction of the consistent texture band and the main direction of the regional wind field is calculated; wherein, the wind erosion direction response parameter It can be represented as: ;in This indicates the directional angle between the direction of the consistent texture band and the main direction of the regional wind field. The smaller the directional angle, the lower the angle. The closer the value is to 1, the more consistent the particle migration direction is with the prevailing wind direction in the region, and the stronger the wind erosion response in the corresponding area. Subsequently, the number of directional texture bands containing particle migration traces per unit area is counted to determine the texture band density; then, the texture band density is combined with... , Amplitude of regional brightness change Passing the exam Assigning values ​​to the activity level of particle migration directions in each direction. Calculate the surface wind erosion intensity variation parameters in various directions. ;in, It can be represented as: ;in , , The values ​​were set to 0.2, 0.4, and 0.4, respectively. The denser the texture bands, the more pronounced the regional brightness changes, and the higher the particle migration activity, the greater the change parameter of the surface wind erosion intensity in the corresponding region. Subsequently, the particle migration direction activity, wind erosion direction response parameters, and surface wind erosion intensity change parameters were weighted and evaluated according to weights of 0.4, 0.3, and 0.3, respectively, and the weighted results were used to screen the particle migration direction of characteristic regions at corresponding time points.

[0073] It is worth noting that the method for determining the dominant wind direction of the corresponding region or feature region of the consistent texture band in each direction includes: firstly, statistically analyzing the directional distribution of the dominant wind direction of the region corresponding to the initial surface region of the consistent texture band or feature region in each direction, and then performing directional clustering analysis on the dominant wind direction of the region between adjacent initial surface regions; wherein, when the angle between the dominant wind directions of adjacent regions is less than 20°, the corresponding regions are classified into the same wind direction category. Subsequently, the frequency of wind direction corresponding to each initial surface region within the same wind direction category is statistically analyzed, and the dominant wind direction of the region with the highest frequency is taken as the dominant wind direction of the region corresponding to the consistent texture band or feature region.

[0074] Of particular importance is the selection of particle migration directions corresponding to feature regions at each time point, including:

[0075] Based on the coverage and density of texture bands with consistent direction at each time node, the extension direction that meets the preset threshold for both coverage and density is selected as the candidate particle migration direction.

[0076] The weighted evaluation results are processed by directional response assignment to be used as the directional response intensity values ​​corresponding to the migration directions of each candidate particle.

[0077] Based on the relationship between the magnitude of the directional response intensity values ​​corresponding to the migration directions of each candidate particle, the migration directions of the candidate particles whose directional response intensity values ​​are greater than the preset response threshold and whose directional response intensity values ​​are the largest in the corresponding feature regions are determined as the migration directions of the particles in the feature regions of the corresponding time nodes.

[0078] In this embodiment, the coverage area of ​​the strips corresponding to the consistent texture bands and the number of texture bands per unit area are first counted at each time point. The strip coverage area is represented by the proportion of the pixel area covered by the consistent texture band to the total area of ​​the corresponding feature region, and the texture band density is represented by the number of texture bands within a 100×100 pixel area. When the strip coverage ratio corresponding to a certain extension direction exceeds 18% and the number of texture bands exceeds 6, this extension direction is selected as a candidate particle migration direction. The regional particle migration activity level, wind erosion direction response parameters, and surface wind erosion intensity change parameters are mapped to the corresponding candidate particle migration directions; wherein, the direction response intensity value... Using the weighted result of parameters in the corresponding direction, the calculation formula can be expressed as: ;in , , The values ​​are 0.4, 0.3, and 0.3, respectively. Then, the directional response intensity values ​​corresponding to the migration directions of each candidate particle are sorted and statistically analyzed; when the directional response intensity value corresponding to the migration direction of a candidate particle is greater than 0.65 and is the maximum value in the corresponding feature region, the migration direction of the candidate particle is determined as the particle migration direction of the feature region of the corresponding time node.

[0079] Optionally, establishing the regional wind interaction relationship in step S3 includes:

[0080] Based on the main direction of the regional wind field and the continuous state of wind propagation corresponding to each characteristic region, wind speed parameters and wind frequency variation parameters corresponding to each characteristic region are extracted. Based on the terrain elevation data, slope parameters, aspect parameters and terrain undulation parameters of each characteristic region are extracted.

[0081] In this embodiment, the average wind speed and the frequency of occurrence of the prevailing wind direction over a continuous 24 hours are first extracted based on the prevailing wind direction corresponding to each characteristic region. The wind speed parameter is represented by the average wind speed value over the continuous time period, and the wind frequency variation parameter is represented by the proportion of the prevailing wind direction in all wind direction records. Subsequently, the slope parameter, aspect parameter, and topographic relief parameter corresponding to each characteristic region are calculated based on the topographic elevation raster in the topographic elevation data. The topographic relief parameter is represented by the difference between the maximum and minimum elevations within the region to reflect the degree of surface undulation.

[0082] Based on the spatial angle relationship between the slope aspect parameter and the main direction of the regional wind field, the windward and leeward areas are divided, and the airflow channel constraint correction is performed on the continuous state of wind propagation in combination with the topographic relief parameter to determine the distribution state of the regional airflow channel.

[0083] In a further embodiment, a spatial angle analysis is performed between the slope aspect parameter and the main direction of the regional wind field. When the angle between the slope aspect and the main direction of the regional wind field is less than 45°, the corresponding area is classified as a windward area; when the angle is greater than 135°, the corresponding area is classified as a leeward area. Then, airflow channel constraint correction is performed on the continuous wind propagation state in conjunction with topographic relief parameters. Specifically, the proportion of the number of connections between adjacent areas satisfying the continuous propagation trend to the total number of connections between adjacent areas is used as the original airflow channel connectivity weight. When the regional elevation undulation is continuous and the elevation difference is less than 10m, the airflow channel connectivity weight of the corresponding area is increased by 5% to 20%, while when the regional elevation difference exceeds 20m, the airflow channel connectivity weight of the corresponding area is decreased by 10% to 30%, thus generating the regional airflow channel distribution state.

[0084] The regional wind field transport intensity parameters are calculated based on wind speed parameters, wind frequency variation parameters, and regional airflow channel distribution. The regional wind field transport intensity parameters are then corrected based on slope parameters and topographic relief parameters to determine the wind propagation attenuation state corresponding to different topographic regions.

[0085] In a further embodiment, the regional wind field transport intensity parameter is calculated based on wind speed parameters, wind frequency variation parameters, and the regional airflow channel distribution status; wherein, the regional wind field transport intensity parameter... Represented as: ;in, This indicates the average wind speed in the region. Indicates the frequency of the prevailing wind direction. This indicates the connectivity weight of the regional airflow channels. Subsequently, the regional wind field transport intensity parameter is attenuated and corrected by combining the slope parameter and the terrain undulation parameter; when the slope exceeds 25° or the regional elevation difference exceeds 20m, the wind field transport intensity of the corresponding region is reduced by 10% to 35% to generate the wind propagation attenuation state corresponding to different terrain regions.

[0086] Based on the wind propagation attenuation state and the regional airflow channel distribution, the regional wind propagation path relationship is constructed, and the wind action correlation path between different regions is established by combining the windward and leeward areas, thereby constructing the regional wind action relationship.

[0087] In a further embodiment, wind propagation connectivity between regions is established based on the wind propagation attenuation state and airflow channel distribution state of each region. Specifically, when the difference in wind propagation attenuation coefficient between adjacent regions is less than 0.2 and the regional airflow channel connectivity weight is greater than 0.6, a wind propagation path relationship is established between the corresponding regions. Then, combining the spatial distribution relationship between the windward and leeward regions, regions with continuous wind propagation directions are path-associated to generate the regional wind action relationship corresponding to the target monitoring area.

[0088] Optionally, the detection of surface particle migration trajectories in step S3 includes:

[0089] The continuity of particle migration direction is detected by the directional offset relationship between corresponding particle migration directions at consecutive time nodes.

[0090] Based on the continuity of particle migration direction, the change of migration expansion range of corresponding particle migration direction within continuous time nodes is statistically analyzed, and the particle migration expansion amount is calculated based on the change of migration expansion range.

[0091] In this embodiment, the directional changes of particle migration directions corresponding to each feature region within consecutive time nodes are first statistically analyzed, and the directional offset angle between adjacent time nodes is calculated. When the directional offset angle between three consecutive time nodes is less than 15°, it is determined that the corresponding region exhibits continuity in particle migration direction. Subsequently, area statistics are performed on the migration expansion range corresponding to the particle migration direction within consecutive time nodes. The migration expansion range is represented by the coverage area of ​​the region corresponding to the particle migration direction, and the difference in coverage area between adjacent time nodes is used as the particle migration expansion amount.

[0092] By combining the wind propagation path direction of the corresponding characteristic area in the regional wind force interaction relationship, the directional deflection difference between the particle migration direction and the regional wind propagation path direction is detected.

[0093] The migration direction of particles with a direction deflection difference greater than a preset deflection threshold is taken as the effective particle migration direction, and the surface particle migration trajectory is generated by combining the particle migration expansion amount corresponding to the effective particle migration direction.

[0094] In a further embodiment, the wind propagation path direction corresponding to each feature area is extracted based on the regional wind force interaction relationship, and the directional deflection difference between the particle migration direction and the wind propagation path direction is calculated; wherein, the directional deflection difference is represented by the absolute value of the angle between the two directions. When the directional deflection difference is less than a preset matching degree threshold of 20°, the corresponding particle migration direction is determined as the effective particle migration direction, indicating that the particle migration direction is consistent with the regional wind propagation path direction; particle migration directions with a directional deflection difference not greater than the preset matching degree threshold of 20° are determined as non-wind erosion interference directions. Then, the surface particle migration trajectory is generated by combining the particle migration expansion amount corresponding to the effective particle migration direction; wherein, when the effective particle migration direction remains continuously stable and the particle migration expansion amount continues to increase, it is determined that there is a continuously enhanced surface particle migration trajectory in the corresponding area, and the particle migration expansion amount is used as the trajectory parameter.

[0095] It is worth noting that the surface particle migration trajectory refers to the spatial motion path formed by temporally associating the migration direction and corresponding expansion amount of particles within the same characteristic area at continuous time nodes. Specifically, it is a continuous displacement path formed by the gradual shift of the position of texture bands with consistent direction in remote sensing images. This path reflects the changes in the spatial migration direction and migration intensity of particles under the action of wind. Among them, when the migration direction of adjacent time nodes maintains a small deviation and the expansion amount shows an increasing change, this continuous shift path is identified as a stable particle migration trajectory, which is used to characterize the spatial evolution process of particles continuously transported downwind from the initial position under the action of regional wind erosion.

[0096] Optionally, identifying areas of wind erosion activity in step S3 includes:

[0097] Based on the surface particle migration trajectory corresponding to each characteristic region, several candidate erosion activity regions are divided, and candidate erosion activity regions with continuous wind propagation relationship are combined with wind action correlation paths as the same candidate erosion activity region.

[0098] In this embodiment, regional clustering analysis is first performed based on the migration trajectories of surface particles corresponding to each characteristic region. Adjacent characteristic regions with a particle migration direction deviation of less than 20° and a consistent trend in particle migration expansion are classified as the same candidate erosion activity region. Subsequently, the wind propagation continuity between candidate erosion activity regions is analyzed by combining wind action correlation paths in the regional wind action relationship. When there is a continuous wind propagation path between adjacent candidate erosion activity regions, and the continuous deflection angle of the path direction is less than 15°, the corresponding candidate erosion activity regions are merged to generate a continuous wind erosion propagation region.

[0099] The duration of wind erosion activity is assessed based on the mean of particle migration and expansion and the mean of the difference in direction deflection in each candidate erosion activity area. Candidate erosion activity areas whose duration exceeds a preset duration threshold are selected as wind erosion activity areas.

[0100] In a further embodiment, the average particle migration and expansion amount and the average directional deflection difference of each candidate erosion activity area are statistically analyzed over consecutive time points. The average particle migration and expansion amount reflects the continuous growth of the particle migration range in the area, and the average directional deflection difference reflects the consistency between the particle migration direction and the wind propagation direction in the area. Then, the average particle migration and expansion amount (weight 0.6) and the average directional deflection difference (weight 0.4) are weighted and calculated to generate a wind erosion activity persistence parameter. When the wind erosion activity persistence parameter exceeds the persistence threshold of 0.65, the corresponding candidate erosion activity area is determined as a wind erosion activity area.

[0101] Optionally, the effective wind erosion area detected in step S4 includes:

[0102] The vegetation coverage of each characteristic region is calculated based on the reflectance of each band in the remote sensing image data, and vegetation degradation areas are divided based on the changes in vegetation coverage.

[0103] In this embodiment, the reflectance of the red and near-infrared bands in the remote sensing image data is first extracted, and the vegetation coverage corresponding to each feature region is calculated based on the reflectance of each band; wherein, the vegetation coverage rate... It can be represented as: ;in, This represents the normalized vegetation index of the corresponding feature region. and These represent the maximum and minimum values ​​of the vegetation index within the target monitoring area, respectively. Subsequently, the changes in vegetation coverage of each characteristic area within consecutive time nodes are statistically analyzed; when the vegetation coverage decreases by more than 10% between two consecutive time nodes, and the continuous decrease lasts for more than 5 days, the corresponding area is classified as a vegetation degradation area.

[0104] The spatial intersection of wind erosion activity areas and vegetation degradation areas is performed, and areas with a spatial overlap ratio greater than a preset overlap threshold are selected as effective wind erosion areas.

[0105] In this embodiment, a spatial overlay analysis is performed on the wind erosion activity area and the vegetation degradation area, and the proportion of the overlapping area to the total area of ​​the wind erosion activity area is calculated; wherein, the spatial overlap ratio is... It can be represented as: ;in, Indicates the area of ​​the intersection region. This represents the total area of ​​wind erosion activity. When the spatial overlap ratio exceeds 0.7, the corresponding intersection area is defined as the effective wind erosion area, indicating that the area simultaneously exhibits continuous wind erosion activity and significant vegetation degradation.

[0106] Of particular importance is that step S4, which divides the effective wind erosion areas into corresponding wind erosion levels, includes:

[0107] The regional wind erosion intensity is obtained by weighting the migration trajectory of surface particles, the decline in vegetation coverage, and the amplitude of regional brightness changes in each effective wind erosion area.

[0108] The intensity of regional wind erosion is compared with a preset erosion level correspondence table to classify wind erosion levels.

[0109] In this embodiment, three basic quantities are first extracted from each effective wind erosion area: surface particle migration trajectory value (obtained by normalization of migration expansion), vegetation cover reduction rate (average of NDVI reduction rate over consecutive time nodes), and regional brightness change amplitude (calculated based on the gray-scale mean difference of multi-temporal images). A vegetation cover reduction rate exceeding 8% is considered a significant degradation contribution, and a brightness change amplitude exceeding 0.12 is considered a high-reflectance disturbance contribution. Subsequently, the three parameters are normalized to a uniform scale, controlling their values ​​between 0 and 1, and then weighted and superimposed with weights of 0.5, 0.3, and 0.2 to obtain the regional wind erosion intensity value. The calculated regional wind erosion intensity is matched and compared with a preset erosion level correspondence table. The correspondence table is divided into four level intervals: light, moderate, heavy, and extremely heavy, according to 0-0.3, 0.3-0.6, 0.6-0.8, and 0.9–1.0. When the regional wind erosion intensity falls into the corresponding interval, the effective wind erosion area is divided into the corresponding wind erosion level, and the spatial distribution result of the corresponding level is output.

[0110] Optionally, this application also provides a soil wind erosion monitoring system 100 based on remote sensing image data, used to execute the soil wind erosion monitoring method based on remote sensing image data as described above. The soil wind erosion monitoring system 100 based on remote sensing image data includes:

[0111] Data acquisition module 101 is used to acquire remote sensing image data, regional wind field data and terrain elevation data corresponding to the target monitoring area;

[0112] The wind erosion feature determination module 102 is used to determine the consistency of the surface texture direction based on the difference in texture gradient change features between corresponding feature regions of adjacent image frames in remote sensing image data; and to identify particle migration traces based on the consistency of the surface texture direction and regional wind field data in order to screen the particle migration direction.

[0113] The activity area identification module 103 is used to establish the relationship between regional wind force based on regional wind field data and topographic elevation data, and to detect the migration trajectory of surface particles in combination with the particle migration direction corresponding to different time nodes, so as to identify the wind erosion activity area.

[0114] The erosion level classification module 104 is used to detect effective wind erosion areas based on the migration trajectory of surface particles and changes in vegetation cover in each wind erosion activity area, and to classify the wind erosion level corresponding to each effective wind erosion area.

[0115] The erosion evolution analysis module 105 is used to generate the spatial distribution results of soil wind erosion in the target monitoring area based on the wind erosion level, and to construct the regional wind erosion evolution sequence by combining the changes in wind erosion level at different time points.

[0116] Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of the equivalents of the application are intended to be included within the invention.

[0117] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features of the invention herein.

Claims

1. A method for monitoring soil wind erosion based on remote sensing image data, characterized in that, Includes the following steps: Step S1: Collect remote sensing image data, regional wind field data, and topographic elevation data corresponding to the target monitoring area; Step S2: Determine the consistency of surface texture direction based on the differences in texture gradient change characteristics between corresponding feature regions of adjacent image frames in the remote sensing image data; identify particle migration traces based on the consistency of surface texture direction and regional wind field data to filter particle migration directions. Step S3: Establish the regional wind force relationship based on regional wind field data and topographic elevation data, and detect the migration trajectory of surface particles in combination with particle migration direction to identify areas of wind erosion activity; Step S4: Based on the migration trajectory of surface particles and changes in vegetation cover in each wind erosion activity area, detect the effective wind erosion area and classify the wind erosion level corresponding to each effective wind erosion area; Step S5: Generate the spatial distribution results of soil wind erosion in the target monitoring area based on the wind erosion level, and construct the regional wind erosion evolution sequence by combining the changes in wind erosion level at different time points.

2. The method for monitoring soil wind erosion based on remote sensing image data according to claim 1, characterized in that, Methods for dividing feature regions include: Extract the surface remote sensing response features corresponding to the target monitoring area from the remote sensing image data, and divide the surface remote sensing response features into several initial surface areas based on the differences in spectral variation and spatial texture distribution. Based on the regional wind field data, the wind field direction of each initial surface area is detected to determine the distribution of wind field direction in each initial surface area; based on the distribution of wind field direction, the main direction of regional wind field in each initial surface area is clustered and statistically analyzed. Based on the topographic elevation data, the slope aspect parameters corresponding to each initial surface area are extracted, and the continuous wind propagation status corresponding to each initial surface area is detected based on the spatial correspondence between the slope aspect parameters and the main direction of the regional wind field, thereby identifying the continuous wind propagation path. If the difference in surface remote sensing response characteristics between adjacent initial surface areas is less than a preset difference threshold, and there is a continuous wind propagation path, then the corresponding adjacent initial surface areas will be merged, and the merged area will be used as the feature area.

3. The method for monitoring soil wind erosion based on remote sensing image data according to claim 1, characterized in that, Step S2, determining the consistency of surface texture direction, includes: Based on the differences in texture gradient change characteristics, determine the direction and intensity of grayscale gradient changes between pixels within a preset neighborhood window. The main gradient direction of the corresponding neighborhood window is determined based on the proportion of gray-level gradient change directions, and the gradient magnitude of the corresponding neighborhood window is calculated based on the intensity of each gradient change. The consistency of texture change direction is detected based on the directional offset between the main gradient directions corresponding to adjacent time nodes, and the texture structure change state is determined by combining the trend of gradient magnitude change. Feature regions whose directional offset is less than a preset directional offset threshold and whose texture structure changes meet the preset directional continuous expansion condition are identified as having consistent surface texture direction.

4. The method for monitoring soil wind erosion based on remote sensing image data according to claim 1, characterized in that, Step S2, which involves identifying particle migration traces, includes: Extract directional consistent texture bands from image frames based on the surface texture direction consistency determination results; Perform position offset detection on the corresponding texture bands with consistent orientation in adjacent image frames, and calculate the spatial displacement direction and displacement distance of each texture band with consistent orientation. If the spatial displacement direction of the texture bands with consistent orientation arranged continuously along similar directions in adjacent image frames satisfies the preset orientation concentration condition, and the displacement distance satisfies the preset increasing change condition within consecutive time nodes, then it is determined that the texture band with consistent orientation has traces of particle migration.

5. The method for monitoring soil wind erosion based on remote sensing image data according to claim 4, characterized in that, After determining that there are traces of particle migration in the corresponding feature region, the following steps are also included: Based on the spectral characteristics of each feature region, determine the brightness variation amplitude of the corresponding texture band with consistent direction in each feature region; Based on the changes in the edge position and coverage of the directional texture bands within consecutive time nodes, the expansion range of the directional texture bands with particle migration traces is determined, and the activity level of the particle migration direction is evaluated based on the expansion range change status. The main direction of the regional wind field at the corresponding time node of the image frame is extracted from the regional wind field data. The wind erosion direction response parameters are calculated based on the degree of direction matching between the extension direction of the consistent texture band and the main wind direction of the regional wind field. The distribution of texture bands with consistent direction of particle migration traces was statistically analyzed to determine the density of texture bands, and the surface wind erosion intensity variation parameters were calculated by combining the amplitude of regional brightness variation and the activity of particle migration direction. The particle migration direction activity level, wind erosion direction response parameters, and surface wind erosion intensity change parameters are weighted and evaluated to screen the particle migration direction corresponding to the characteristic areas at each time point.

6. The method for monitoring soil wind erosion based on remote sensing image data according to claim 1, characterized in that, Step S3, establishing the regional wind force interaction relationship, includes: Based on the main direction of the regional wind field and the continuous state of wind propagation corresponding to each characteristic region, wind speed parameters and wind frequency variation parameters corresponding to each characteristic region are extracted. Based on the terrain elevation data, slope parameters, aspect parameters and terrain undulation parameters of each characteristic region are extracted. Based on the spatial angle relationship between the slope aspect parameter and the main direction of the regional wind field, the windward and leeward areas are divided, and the airflow channel constraint correction is performed on the continuous state of wind propagation in combination with the topographic relief parameter to determine the distribution state of the regional airflow channel. The regional wind field transport intensity parameters are calculated based on wind speed parameters, wind frequency variation parameters, and regional airflow channel distribution. The regional wind field transport intensity parameters are then corrected based on slope parameters and topographic relief parameters to determine the wind propagation attenuation state corresponding to different topographic regions. Based on the wind propagation attenuation state and the regional airflow channel distribution, the regional wind propagation path relationship is constructed, and the wind action correlation path between different regions is established by combining the windward and leeward areas, thereby constructing the regional wind action relationship.

7. The method for monitoring soil wind erosion based on remote sensing image data according to claim 1, characterized in that, Step S3, detecting the migration trajectory of surface particles, includes: The continuity of particle migration direction is detected by the directional offset relationship between corresponding particle migration directions at consecutive time nodes. Based on the continuity of particle migration direction, the change of migration expansion range of corresponding particle migration direction within continuous time nodes is statistically analyzed, and the particle migration expansion amount is calculated based on the change of migration expansion range. By combining the wind propagation path direction of the corresponding characteristic area in the regional wind force interaction relationship, the directional deflection difference between the particle migration direction and the regional wind propagation path direction is detected. The migration direction of particles with a direction deflection difference greater than a preset deflection threshold is taken as the effective particle migration direction, and the surface particle migration trajectory is generated by combining the particle migration expansion amount corresponding to the effective particle migration direction.

8. The method for monitoring soil wind erosion based on remote sensing image data according to claim 1, characterized in that, Step S3, which identifies areas of wind erosion activity, includes: Based on the surface particle migration trajectory corresponding to each characteristic region, several candidate erosion activity regions are divided, and candidate erosion activity regions with continuous wind propagation relationship are combined with wind action correlation paths as the same candidate erosion activity region. The duration of wind erosion activity is assessed based on the mean of particle migration and expansion and the mean of the difference in direction deflection in each candidate erosion activity area. Candidate erosion activity areas whose duration exceeds a preset duration threshold are selected as wind erosion activity areas.

9. The method for monitoring soil wind erosion based on remote sensing image data according to claim 1, characterized in that, The effective wind erosion area detected in step S4 includes: The vegetation coverage of each characteristic region is calculated based on the reflectance of each band in the remote sensing image data, and vegetation degradation areas are divided based on the changes in vegetation coverage. The spatial intersection of wind erosion activity areas and vegetation degradation areas is performed, and areas with a spatial overlap ratio greater than a preset overlap threshold are selected as effective wind erosion areas.

10. A soil wind erosion monitoring system based on remote sensing image data, characterized in that, For performing the soil wind erosion monitoring method based on remote sensing image data as described in claim 1, the soil wind erosion monitoring system based on remote sensing image data includes: The data acquisition module is used to collect remote sensing image data, regional wind field data, and terrain elevation data corresponding to the target monitoring area; The wind erosion feature determination module is used to determine the consistency of surface texture direction based on the differences in texture gradient change features between corresponding feature regions of adjacent image frames in remote sensing image data; and to identify particle migration traces based on the consistency of surface texture direction and regional wind field data in order to filter particle migration direction. The activity area identification module is used to establish the relationship between regional wind forces based on regional wind field data and topographic elevation data, and to detect the migration trajectory of surface particles by combining the particle migration direction corresponding to different time nodes, so as to identify the wind erosion activity area. The erosion level classification module is used to detect effective wind erosion areas based on the migration trajectory of surface particles and changes in vegetation cover in each wind erosion activity area, and to classify the wind erosion level corresponding to each effective wind erosion area. The erosion evolution analysis module is used to generate the spatial distribution results of soil wind erosion in the target monitoring area based on the wind erosion level, and to construct the regional wind erosion evolution sequence by combining the changes in wind erosion level at different time points.

Images