Ecological hydrological element sky-ground integrated remote sensing monitoring method and system

Abstract

This invention discloses an integrated space-air-ground remote sensing monitoring method and system for eco-hydrological elements, belonging to the field of remote sensing monitoring technology. The method includes: acquiring a slope aspect layer and comparing slope abrupt changes; extracting fracture boundary segments; grouping directional changes to screen stable paths; extracting humidity cycle markers for abrupt changes; tracking NDVI growth paths to generate response sequences; mapping and reconnecting layer boundaries to form an updated path set. This invention enhances the spatial perception of topographic boundary changes by guiding slope abrupt changes through the dominant runoff direction; groups paths according to directional vector change trends to improve the continuity and stability of boundary extraction; combines trend offset blocks in the humidity cycle for time annotation to strengthen temporal comparison during the humidity response process; constructs vegetation response chains by continuously splicing NDVI growth paths; and uses path mapping to drive dynamic adjustments to the remote sensing boundary structure, making the boundary update process more closely aligned with the spatiotemporal progression characteristics of ecological disturbances.

Description

Technical Field

[0001] This invention relates to the field of remote sensing monitoring technology, and in particular to an integrated space-air-ground remote sensing monitoring method and system for eco-hydrological elements. Background Technology

[0002] The field of remote sensing monitoring technology encompasses the acquisition, processing, and analysis of Earth observation information. Its core components include acquiring reflection information from different electromagnetic bands based on space, airborne, and ground platforms to obtain multi-dimensional information such as regional land cover, ecological environment, meteorology, and hydrology. Through radiometric and geometric correction, feature extraction, and spatial analysis of multi-source remote sensing images, it enables the identification and tracking of changes in land surface elements. This technology system integrates optical remote sensing, radar, thermal infrared remote sensing, and lidar, possessing characteristics such as wide coverage, high temporal resolution, and strong dynamic observation capabilities. It is widely used in fields such as land use change, ecological environment assessment, disaster monitoring, agricultural assessment, and hydrological monitoring.

[0003] The integrated air-ground remote sensing monitoring method for eco-hydrological elements refers to the use of satellite remote sensing, aerial remote sensing, and ground-based measurements to acquire multi-scale observation data of core eco-hydrological elements such as precipitation, soil moisture, surface runoff, evapotranspiration, and vegetation cover for regional ecological and hydrological process monitoring tasks. This type of method typically extracts vegetation indices and high-resolution aerial images of water body boundaries from satellite multispectral remote sensing images to supplement surface detail information. Ground instruments collect measured data on precipitation, runoff, and soil moisture. Based on the temporal matching and spatial positioning of remote sensing images, data fusion and supplementation are performed to construct a spatially continuous eco-hydrological element dataset to complete the comprehensive monitoring of eco-hydrological processes in the target area.

[0004] Existing technologies rely heavily on rule templates or single-point thresholds for spatial boundary identification, lacking a structural response mechanism based on directional jumps and upstream / downstream extension logic. This can easily lead to problems such as ambiguous location of fracture features. In monitoring humidity and vegetation changes, the focus is on comparative analysis of overall trends, neglecting the tracking dimension of turning points within the cycle. This makes it difficult to accurately reveal the temporal and locational characteristics of the generation and expansion of local disturbances. In terms of layer updates, the technology has failed to establish a mapping relationship from ecological response paths to boundary adjustments, resulting in remote sensing boundaries failing to reflect actual ecological process changes. This leads to problems such as response lag, path breaks, and spatial disconnect. Summary of the Invention

[0005] To address the technical problems existing in the prior art, this invention provides an integrated space-air-ground remote sensing monitoring method and system for eco-hydrological elements, the specific technical solution of which is as follows:

[0006] On the one hand, an integrated space-air-ground remote sensing monitoring method for eco-hydrological elements is provided, including the following steps:

[0007] S1: Obtain the surface slope aspect layer, select the dominant runoff direction as the reference line, compare the slope values ​​of adjacent pixels along the direction, extract continuous jump points, extend the upstream and downstream range, record the abrupt change location, and obtain the candidate region fragment of the fault boundary.

[0008] S2: Based on the candidate region segments of the fracture boundary, extract the line segment direction vectors, sort the direction change values, group them according to the change trend, filter the continuous boundary lines with consistent direction frequency repetition, retain the corresponding pixels, and obtain a set of stable boundary path segments.

[0009] S3: Based on the set of stable boundary path segments, extract the soil moisture periodic sequence within the region, locate records of unrecovered trends, and combine the continuous stagnant blocks marked on the time axis to obtain a labeling layer of moisture response jump areas;

[0010] S4: Based on the labeled layer of the humidity response jump area, extract the NDVI amplitude rise segment periodically, track the NDVI growth path in the adjacent direction, arrange the path segments according to time, and obtain the vegetation main response path sequence group.

[0011] S5: Based on the vegetation main response path sequence group, map the path edge to the remote sensing layer boundary, replace the original line segment according to the path and reconnect the cell boundary, and connect them to the whole map in sequence to obtain the remote sensing layer boundary update path set.

[0012] As a further aspect of the present invention, the candidate region segment of the fracture boundary includes pixels with abrupt changes in slope direction, a group of continuously jumping nodes, and upstream and downstream connected pixel blocks; the set of stable boundary path segments includes a sequence of direction vector arrangements, hierarchical grouping of direction changes, and continuously distributed boundary path segments; the labeling layer of the humidity response jump region includes soil humidity time trajectory, humidity trend turning point, trend offset patch area, and time series label; the vegetation main response path sequence group includes NDVI amplitude rising segment, continuously growing path chain, neighborhood extension path segment, and time series connection channel line; and the set of remote sensing layer boundary update paths includes boundary mapping reference points, path position replacement nodes, sequential reconnection boundary chain segments, and layer boundary connection trajectory.

[0013] As a further embodiment of the present invention, the continuous path boundary line refers to the pixel path line segments that have the same directional change trend, the same directional frequency, and are spatially continuous in the candidate region of the fracture boundary.

[0014] The continuous stagnant block refers to a pixel region segment in the humidity sequence where the direction of change has not been restored and the region remains stagnant and without reversal on the time axis.

[0015] As a further aspect of the present invention, the trend failure record refers to the time record point in the humidity cycle sequence where the pixel does not return to the original trend direction after experiencing a trend change.

[0016] The pixel boundary refers to the spatial boundary line formed by connecting adjacent pixels in the remote sensing layer.

[0017] As a further aspect of the present invention, the specific steps of S1 are as follows:

[0018] S101: Obtain the surface slope aspect data layer in the ecological remote sensing area, select the dominant direction of surface runoff as the reference line, extract the slope direction data of adjacent pixels sequentially along the reference line, compare the slope direction of the pixel with the search direction of the previous pixel, record the position sequence of the direction jump, and obtain the slope aspect jump position record group.

[0019] S102: Based on the slope aspect change location record group, expand the adjacent pixel data in the three rows of the upstream and downstream of the extended region, retrieve the directional position of the continuous connection of pixels in space, mark the upstream and downstream extension zones according to the spatial continuity relationship, and obtain the slope aspect change expansion region zone.

[0020] S103: Based on the slope aspect change extension zone, sequentially read the cell number and coordinates at the slope aspect change position, summarize the spatially continuous cell combinations with jump characteristics, output the number list in the recording order, and obtain the candidate region fragment of the fracture boundary.

[0021] As a further aspect of the present invention, the specific steps of S2 are as follows:

[0022] S201: Based on the boundary line segments in the candidate region of the fracture boundary, obtain the direction vector value corresponding to the line segment, connect adjacent vectors according to the line segment arrangement order, remove line segment combinations with discontinuous direction changes, and obtain a sequence of direction change values.

[0023] S202: Based on the change values ​​in the numerical sequence of direction changes, calculate the rate of change of each segment of direction change difference, divide the direction change interval, and group the paths whose rate values ​​fall within the same range into the same group to obtain the direction change rate interval.

[0024] S203: Based on the spatial distribution of path segments in the range of directional change rates, filter out pixel segments that are adjacent in position and whose directional trend is not interrupted, exclude discontinuous path positions, and obtain a set of stable boundary path segments.

[0025] As a further aspect of the present invention, the specific steps of S3 are as follows:

[0026] S301: Based on the region corresponding to the set of stable boundary path segments, collect the soil moisture index of the pixels within the period, arrange the moisture data at the same position in chronological order, and distinguish the storage sequence trajectory according to the pixel number to obtain the soil moisture periodic sequence group.

[0027] S302: Based on the humidity sequence corresponding to the pixel in the soil moisture periodic sequence group, identify the segments in the sequence with continuous and consistent change direction, find the fluctuation interruption point in the continuous segment and record the position change trend, calculate the reversal delay value between the change amplitude and the trend direction in each segment of the sequence, and obtain the humidity change delay trend value.

[0028] S303: Based on the fact that there are change segments in the humidity change delay trend value where the trend has not returned, filter out the pixel areas with consecutive numbers, synchronously mark the time coordinates of the corresponding blocks, add jump trend direction information, and obtain the humidity response jump area annotation layer.

[0029] As a further aspect of the present invention, the specific steps of S4 are as follows:

[0030] S401: Based on the location of the region in the humidity response jump region labeling layer, read the NDVI value of the corresponding pixel from the remote sensing layer periodically, arrange the periodic records in chronological order, divide the continuous NDVI amplitude sequence by region, and obtain the NDVI time series amplitude group.

[0031] S402: Call the sequence content in the NDVI time series amplitude group, identify the changing segments with continuous growth characteristics, connect the position chains with the same direction of change, mark the start and end index positions between continuous segments, and obtain the NDVI growth trend path segments.

[0032] S403: Based on the time sequence of the NDVI growth trend path segments, the continuous path segments are connected in sequence, and the path structure in the growth direction is extended to the adjacent area. The connecting positions and sequence numbers are connected in series to obtain the vegetation main response path sequence group.

[0033] As a further aspect of the present invention, the specific steps of S5 are as follows:

[0034] S501: Based on the path edge position in the vegetation main response path sequence group, mark the corresponding coordinates of the path endpoints and turning points in the remote sensing layer, match the coordinates with the node positions on the original boundary line, and obtain the path boundary mapping corresponding set.

[0035] S502: Based on the node order in the path boundary mapping corresponding set, compare the position sequence of the path line with the original boundary line, replace the position content of the corresponding line segment on the original boundary line with the path node, synchronize the position order of the path line with the original boundary arrangement, and obtain the boundary path replacement sequence.

[0036] S503: Based on the line segment order in the boundary path replacement sequence, organize each boundary segment according to the connection order of the path nodes, maintain the continuity of the layer number between the line segments, and connect each boundary path segment in sequence to obtain the remote sensing layer boundary update path set.

[0037] On the other hand, an integrated space-air-ground remote sensing monitoring system for eco-hydrological elements is provided, including:

[0038] The slope aspect analysis module acquires the surface slope aspect data layer of the ecological remote sensing area, selects the dominant direction of surface runoff as the reference line, compares the slope direction of adjacent pixels along the reference direction, identifies continuous jump positions, and expands to the upstream and downstream directions in combination with the spatial connection relationship of pixels. Gradually, it collects pixels with abrupt slope changes and connection nodes to obtain candidate area fragments of fracture boundaries.

[0039] The fracture extraction module extracts internal boundary line segments based on the candidate fracture boundary region fragments, obtains direction vectors, arranges the direction change sequence, groups continuous vectors according to the change order, and filters boundary path segments with consistent direction change frequency and continuous distribution to obtain a set of stable boundary path segments.

[0040] The humidity response module, based on the set of stable boundary path segments, extracts the soil moisture index within the path area period, locates the humidity change trajectory of the pixel time series, identifies segments with continuous and consistent directions, retrieves the position of stagnant change, locates the area block where the trend has not returned to the initial state, and obtains the humidity response jump area annotation layer.

[0041] The vegetation path module extracts the NDVI amplitude change information of the region periodically based on the humidity response jump area annotation layer, filters the continuously rising path segments according to the pixel order, connects adjacent rising paths and extends along the adjacent direction, and connects the path segments in time order to obtain the vegetation main response path sequence group.

[0042] The boundary reconstruction module maps the path edge positions to the remote sensing layer boundary lines based on the vegetation main response path sequence group, matches the path lines with the boundary lines in sequence, replaces the boundary line segments one by one, and connects the boundary segments along the path nodes in sequence to obtain the remote sensing layer boundary update path set.

[0043] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0044] In this invention, slope abrupt changes are extracted by guiding the dominant runoff direction, enhancing the spatial perception of the location of terrain boundary changes. Paths are grouped according to the trend of directional vector changes to improve the continuity and stability of boundary extraction. Time labeling is performed by combining trend offset blocks in the humidity cycle to strengthen the temporal comparison in the humidity response process. Vegetation response chains are constructed by continuously splicing the NDVI growth path. Path mapping promotes the dynamic adjustment of the remote sensing boundary structure, making the boundary update process more in line with the spatiotemporal progression characteristics of ecological disturbance. Attached Figure Description

[0045] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0046] Figure 1 This is a schematic diagram of the steps of the present invention;

[0047] Figure 2 This is a system module diagram of the present invention. Detailed Implementation

[0048] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0049] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0050] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning.

[0051] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0052] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0053] Please see Figure 1 This invention provides an integrated space-air-ground remote sensing monitoring method for eco-hydrological elements, comprising the following steps:

[0054] S1: Obtain the surface slope aspect data layer in the ecological remote sensing area, select the dominant direction of surface runoff as the reference line, compare the slope direction of adjacent pixels in sequence along the direction, extract the locations where the slope changes continuously according to spatial continuity, extend the range of adjacent pixels along the upstream and downstream directions, and record the set of pixels where the slope direction changes abruptly in sequence to obtain the candidate area fragment of the fault boundary.

[0055] S2: Based on the boundary line segments in the candidate region of the fracture boundary, extract the corresponding direction vector value of each segment, arrange the related direction change values ​​in sequence, group them in order of increasing direction change degree, select the boundary path with consistent direction change frequency and continuous distribution of segments as the current reference baseline, retain the continuous pixel sequence in the corresponding path, and obtain the set of stable boundary path segments.

[0056] S3: Based on the region corresponding to the set of stable boundary path segments, the soil moisture index is extracted in each cycle in sequence. The cycle sequence is located according to the cell number. Stagnation points are retrieved in the records with continuous and consistent moisture change direction. The location blocks that have not returned to the initial trend in continuous change are located. Combined with the synchronous annotation of the time axis, the moisture response jump area annotation layer is obtained.

[0057] S4: Based on the humidity response jump area annotation layer, each region is labeled, and the NDVI time series amplitude is obtained from the remote sensing layer periodically. The rising segments of NDVI amplitude change are extracted point by point, the paths of continuous growth trend are connected, and the path segments of continuous NDVI value growth in the adjacent direction are tracked and linked into channel paths in chronological order to obtain the vegetation main response path sequence group.

[0058] S5: Based on the path edge position in the vegetation main response path sequence group, map it to the original boundary line in the current remote sensing layer. Compare the position order of the path line and the original boundary line, call the current node path to replace the position, reconnect the line segments according to the path order, and connect each boundary segment of the current layer continuously to obtain the remote sensing layer boundary update path set.

[0059] Among them, the candidate region fragments of the fracture boundary include pixels with abrupt changes in slope direction, groups of nodes with continuous jumps, and upstream and downstream connected pixel blocks; the set of stable boundary path segments includes the sequence of direction vector arrangements, hierarchical grouping of direction changes, and continuously distributed boundary path segments; the labeling layer of humidity response jump regions includes soil humidity time trajectory, humidity trend turning point, trend offset patch area, and time series label; the vegetation main response path sequence group includes NDVI amplitude rising segment, continuous growth path chain, neighborhood extension path segment, and time series connection channel line; and the set of remote sensing layer boundary update paths includes boundary mapping reference point, path position replacement node, sequential reconnection boundary chain segment, and layer boundary connection trajectory.

[0060] The specific steps of S1 are as follows:

[0061] S101: Obtain the surface slope aspect data layer in the ecological remote sensing area, select the dominant direction of surface runoff as the reference line, extract the slope direction data of adjacent pixels sequentially along the reference line, compare the slope direction of the pixel with the search direction of the previous pixel, record the position sequence of the direction jump, and obtain the slope aspect jump position record group.

[0062] First, high-resolution remote sensing imagery data is selected within the actual area, and slope aspect values ​​are extracted based on this data. The extraction process must be performed on a per-pixel basis. The slope aspect value of each pixel is determined by the elevation changes around it. For example, if the elevation of a pixel increases relative to its eastern neighbor, the slope aspect of that pixel points eastward. If adjacent pixels have the same elevation, their slope aspect remains horizontal. For a watershed area, the main runoff direction can be extracted using a topographic map. Assuming this direction is southwest, the slope aspect values ​​of adjacent pixels along the southwest direction in the remote sensing data must be extracted row by row, and the slope aspect angle of each pixel must be recorded. For example, if the slope aspect of a pixel is 45 degrees, and the slope aspect angle of the previous pixel is 90 degrees, then the angle change is 45 degrees. This operation is performed continuously to statistically analyze the slope aspect angle differences of all adjacent pixels along the entire direction. To determine whether a location represents a jump, a reference value for the angle jump must be established, referencing the variation in slope aspect angles. For example, 45 degrees is considered the standard interval angle. Whenever the directional change between adjacent pixels equals or exceeds this interval angle, it is recorded as a jump point. Taking a set of actual data as an example, if the slope aspect values ​​are 90 degrees, 135 degrees, 180 degrees, 225 degrees, and 90 degrees, then the jump between 225 degrees and 90 degrees is a 135-degree jump, and this location should be marked. The recorded jump location must be combined with its row and column number and spatial location identifier in the remote sensing data. Subsequently, at the identified jump point, the data of adjacent pixels in both upstream and downstream directions are extended along the runoff direction. The extension range can be defined as a five-pixel range within the two-pixel interval. The continuation of jump points within this range is then assessed. If two or more consecutive jump points appear, they are included in the subsequent recording segment and numbered according to their spatial relative position and number, forming a continuous set of pixels. This ultimately yields the slope aspect jump location record group.

[0063] S102: Based on the slope aspect change location record group, the adjacent pixel data in the three rows upstream and downstream of the extended region are expanded. The directional positions of the pixels that are continuously connected in space are retrieved. The upstream and downstream extension zones are marked according to the spatial continuity relationship to obtain the slope aspect change expansion region zone.

[0064] First, the row and column information of the abruptly changing pixels in the remote sensing layer is retrieved. Based on the current pixel's row, three rows of pixel data are extended upwards and downwards, forming a six-row extension area. Within each extension row, the slope direction values ​​of all adjacent pixels are extracted along the column direction. The difference in slope direction values ​​between adjacent pixels is then sequentially evaluated. If a direction maintains an angle offset of less than 45 degrees from the previous pixel's direction, its position is considered related. When the direction change values ​​are nearly continuous, the pixel is marked as part of a continuously connected directional path. For example, if a group of pixels has directions of 135 degrees, 140 degrees, 142 degrees, 139 degrees, and 141 degrees, it can be included in a continuously connected path. If a direction suddenly becomes 90 degrees, it is considered the end point of the path. This operation needs to be performed on all row and column combinations within the extension area to obtain all possible paths. Subsequently, the pixels along these path directions are numbered according to their spatial location to identify whether they have spatial continuity in the same extension direction. If a pixel in an adjacent path has an angle of less than 30 degrees with the main path direction and is no more than two pixel units away from the preceding and following pixels on the main path, it is marked as part of the extension zone. Based on this, combined with the position sequence of all path pixels in the aforementioned extended rows, a path sequence is constructed according to the pixel number and row / column coordinates. After filtering based on directional consistency and spatial continuity, the extension sequence is retained. Taking a regional example, if the pixel position from R21C10 to R24C10 in the extended region forms a north-south path, and its adjacent pixels from R24C10 to R27C12 are in a direction 15 degrees east of south, then this segment can be included in the extension zone. Finally, all pixel sequences that meet the conditions are integrated into upstream and downstream extension zones and correspondingly marked on the remote sensing layer to obtain the slope aspect abrupt change extension zone.

[0065] S103: Based on the slope aspect abrupt extension zone, the number and coordinates of the pixels at the slope aspect change position are read sequentially, the combination of pixels that are spatially continuous and have abrupt change characteristics is summarized, and the number list is output in the order of recording to obtain the candidate region fragment of the fracture boundary.

[0066] First, the pixel IDs and coordinates in the remote sensing layer are read. The IDs can be constructed by combining the layer row and column numbers with the pixel index. For example, the pixel in row 22, column 10 can be recorded as R22C10, and its ID value in the data table is recorded as P2210. Simultaneously, the slope direction value of this pixel is extracted as an auxiliary identifier. The reading operation needs to traverse all labeled pixels within the entire extended area band, recording their IDs, coordinates, and direction values ​​in a table file for later retrieval. After recording, the pixel IDs need to be arranged in a top-down, left-to-right order according to row and column coordinates, forming a continuous sequence. For example, if pixels R22C10, R23C10, and R24C10 exist, their IDs P2210, P2310, and P2410 are output sequentially. Subsequently, the slope direction of adjacent pixels in the sequence is compared. If the direction value jump exceeds the set deflection reference range (e.g., the direction change exceeds 45 degrees), the jump point is determined to be a location with a sharp change in slope in space. Further checks are performed to see if the two adjacent directions above and below have continuity. If sequences such as 135 degrees, 90 degrees, and 140 degrees exist, the group is determined to be a slope direction jump group and must be included in the recording range. For all jump groups that meet this condition, the continuously distributed pixel combinations in their area are extracted, and it is checked whether at least three consecutive points form a path. If so, it is marked as a continuous jump path segment, and its path number is marked on the layer. All pixel numbers in the segment are re-output according to the path number order, and their corresponding coordinate information is recorded to form a unified segment list. For example, if a segment in the area consists of P2210, P2310, and P2410, and the direction jump value is greater than the reference value, and the three are a continuous north-south line segment in space, then this combination is considered a valid segment and added to the segment list. After completing the list of all pixel segments, the corresponding number list for each jump path segment is output by combining the jump feature judgment conditions and the continuous spatial distribution pattern, thus obtaining the candidate region segment of the fracture boundary.

[0067] The specific steps of S2 are as follows:

[0068] S201: Based on the boundary line segments in the candidate region of the fracture boundary, obtain the direction vector value corresponding to the line segment, connect adjacent vectors according to the arrangement order of the line segments, remove the line segment combinations with discontinuous direction changes, and obtain the direction change numerical sequence.

[0069] First, the coordinates of the two endpoints of each boundary line segment are read. A two-dimensional coordinate difference method is used to calculate the difference between the x and y coordinates of the starting and ending points of each line segment, thereby calculating the direction vector component value of each line segment. For example, when the starting point coordinates of a line segment are (8, 4) and the ending point is (11, 9), its direction vector is (3, 5). Then, all direction vectors are arranged sequentially according to their numbers, and the directions of continuous line segments are connected. During this process, it is necessary to determine whether the angle between two adjacent vectors is within a uniform range of change. If the direction change exceeds the set continuous change range... If the dynamic range threshold, such as the directional difference angle exceeding 25 degrees, is set, the line segment will be removed from the continuous sequence. This threshold needs to be determined based on the characteristics of continuous directional change in the actual remote sensing area. In the gully landform of the Loess Plateau, a value between 15 and 25 degrees is often used. Then, all vector combinations with stable directional changes are retained in sequence. The results of the removal and retention operations are then reviewed to ensure that there are no skip connections in the selected continuous line segment combinations. Finally, the directional vector values ​​corresponding to each line segment in the sequence are summarized, and the set of vector differences corresponding to continuous directional changes is output to obtain the directional change numerical sequence.

[0070] S202: Based on the change values ​​in the numerical sequence of direction changes, calculate the rate of change of each segment of direction change difference, divide the direction change interval, and group the paths whose rate values ​​fall within the same range into the same group to obtain the direction change rate interval.

[0071] First, the angles of adjacent directions are read and their differences are recorded one by one. Negative differences are rewritten as corresponding positive differences, forming a sequence of positive changes. For example, if the angles of two adjacent points are 10, 18, 30, 25, and 40, the changes are recorded as 8, 12, 5, and 15, respectively, and all are written into the change value sequence. In the path record of outdoor moving objects, continuous changes are used to form an execution unit in a fixed number, for example, 10 changes per segment. All changes in the first segment are read, and each change is added together to obtain the overall change description of the segment. For example, if the 10 changes in the first segment are 8, 12, 5, 15, 10, 6, and 9, respectively. If the sequence is 11, 7, 7, then it can be described as a segment with a cumulative change of 90. Then, the corresponding sampling number for this segment is 10 sampling points. Based on this, the speed of this segment is described as a ratio of 90 to 10 sampling points. Without using a formula, this description can be directly recorded as 9 and used as the speed value of this segment. The same process is applied to the second and third segments, forming a sequence of directional speed changes. For example, when the cumulative change of the second segment is 25 corresponding to 10 sampling points, it can be recorded as a speed description value of 2.x; when the cumulative change of the third segment is 50 corresponding to 10 sampling points, it can be recorded as a speed description value of 5.x. By reading each segment sequentially, a complete speed sequence is formed, for example, 9, 2.x. The speed values ​​are then divided into intervals, such as 5.something, 7.something, etc. In the application scenario, the minimum and maximum values ​​of all speed values ​​can be read; for example, the minimum is approximately 2.something and the maximum is approximately 9. This defines the reference boundaries for the speed intervals: 0 to 3 is set as the low-speed interval, 3 to 6 as the medium-speed interval, and 6 to 9 as the high-speed interval. The interval boundaries can be set based on the sorting results of all speed values. For example, speeds in the top 30 percentile after sorting generally do not exceed 3, so setting 3 as the upper limit of the low-speed interval is reasonable. Speeds in the middle 30th to 60th percentiles after sorting are mostly concentrated between 3 and 6, so 3 to 6 is set as the medium-speed interval. The remaining... The speeds are generally concentrated between 6 and 9, so they can be considered as the high-speed range. Then, each speed segment is determined to be within its own range. For example, if the first speed segment is 9, it falls directly into the high-speed range; if the second speed segment is 2.something, it falls into the low-speed range; if the third speed segment is 5.something, it falls into the medium-speed range; and if the fourth speed segment is 7.something, it falls into the high-speed range. The paths are then grouped according to the range number. For example, if all the speed segments of path A are 4.something, 5.something, and 5.something, they all fall into the medium-speed range, then path A is grouped into the medium-speed group. If all the speed segments of path B are 7.something, 8.something, and 9, they all fall into the high-speed range, then path B is grouped into the high-speed group. Finally, the range of directional change rate is obtained.

[0072] S203: Based on the spatial distribution of path segments in the direction change rate interval, filter out cell segments that are adjacent in position and whose directional trend is not interrupted, exclude discontinuous path positions, and obtain a set of stable boundary path segments.

[0073] First, the direction change rate value generated in the previous stage is read. This value is then arranged in pixel order and mapped to the position of each line segment in the raster path layer. The start and end coordinates of each path segment are extracted. The spatial direction continuity is calculated by connecting adjacent path segments. If the end point and start point of two path segments are on the same grid direction channel, such as up / down, left / right, or diagonal eight-neighbor direction, they are considered to be continuous paths in space. Then, the path segments with continuous direction are combined and traversed. The pixel number sequence in each path combination is extracted according to the arrangement order, and it is determined whether there is horizontal offset, vertical discontinuity, or intersection in the path layer. In cases of spatial structure breakage, if the continuous trend of the path within the combined segment is not interrupted, it is marked as a retained state; otherwise, combinations with break points between paths are removed. Further, based on the spatial distribution relationship of the retained path segments, pixel segment groups that satisfy the positional adjacency relationship are retrieved, and it is ensured that the connection relationship of all pixels exists continuously in its original path numbering order. For example, if the path segment number is P21 to P28, the segment group numbering sequence must continuously cover this numbering interval. Segment combinations with skipped numbers or missing connections are excluded. Finally, all continuously numbered segments are summarized, and their corresponding spatial line segment sequences in the path layer are output to obtain the set of stable boundary path segments.

[0074] The specific steps for S3 are as follows:

[0075] S301: Based on the region corresponding to the set of stable boundary path segments, collect the soil moisture index of the pixels within the period, arrange the moisture data at the same position in chronological order, and distinguish the storage sequence trajectory according to the pixel number to obtain the soil moisture periodic sequence group.

[0076] First, the location of the stable boundary path corresponding to the pixel is retrieved in the remote sensing layer. Then, the soil moisture index value of that pixel at different time points is extracted from the time-series remote sensing image. The moisture value collected at each time point must correspond to a specific date to ensure the uniqueness and consistency of the records at each time node. Next, the time points are sorted at fixed intervals to construct a set of moisture value vectors arranged along the time axis. For pixel position (i, j), the constructed moisture sequence should completely record the moisture change data from time period T1 to Tn. Then, during storage, the above vector set is used as the pixel number as the key to establish an independent sequence record table, and each time value corresponding to that pixel is recorded in the table. The interval sequences are stored as independent arrays. In the example, if there are 180 cell numbers in a stable path segment, each number needs to correspond to a complete humidity sequence. The sequence corresponding to a certain number may be: [0.21, 0.24, 0.19, 0.27, 0.31, 0.28], which represents the change in soil moisture in six consecutive remote sensing records. If a cell is found to have missing values ​​in certain time periods, it is replaced by filling in the missing values ​​from the past or taking the average of adjacent dates to ensure the integrity and comparability of the sequence. Finally, the cells of all stable boundary path segments are collected one by one, numbered and archived, and sorted in chronological order to obtain the soil moisture periodic sequence group.

[0077] S302: Based on the humidity sequence corresponding to the pixel in the soil moisture periodic sequence group, identify the segments in the sequence with continuous and consistent change direction, find the fluctuation interruption point in the continuous segment and record the position change trend, calculate the reversal delay value between the change amplitude and the trend direction in each segment of the sequence, and obtain the humidity change delay trend value.

[0078] First, the humidity value is extracted for each pixel and arranged in chronological order. Adjacent humidity values ​​are compared to determine the direction of change. A higher humidity value is recorded as an upward trend, and a lower value as a downward trend. Consecutive values ​​in the same direction are recorded as a single continuous segment. For example, if the humidity sequence for a pixel is 20, 23, 25, 22, 19, 18, 21, 24, then the first three values ​​(22, 19, 18) represent an upward trend and form the first continuous segment. The next three values ​​(22, 19, 18) represent a downward trend and form the second continuous segment. The next continuous segment, consisting of 18 to 21 and 21 to 24, forms the third continuous segment, with the upward trend. Within each continuous segment, the difference between adjacent humidity values ​​is examined one by one. When the difference in magnitude reverses from the previous difference, it is identified as a fluctuation interruption point. For example, in the first continuous segment, the difference between 20 and 23 is 3, and the difference between 23 and 25 is 2. When the difference changes from 3 to 2, the difference value shows a decreasing trend. If the difference then increases from 2 to 3, it is considered a fluctuation interruption point. The sequence index corresponding to this position can be recorded, and its position trend can be marked as changing from decreasing to increasing. To increase the accuracy, this process is performed sequentially on all continuous segments, recording the positional changes of all fluctuation interruption points within each segment. Then, the directional trend of the first and last humidity values ​​within each continuous segment is obtained. For example, if the overall trend of the first continuous segment is upward, the positional trend of each fluctuation interruption point in that segment is read. When a positional trend diverges from the overall trend of the segment, the point where the overall trend begins to diverge from the interruption point is considered the reversal point. The distance between the reversal point and the interruption point is then described. The number of sequence intervals between the two can be directly taken. For example, if the first position of the segment is index 1, the interruption point is index 3, and the reversal point is determined at index 4, then the reversal delay value can be recorded as the interval 1 between 4 and 3. Similarly, the reversal delay description of each continuous segment can be obtained segment by segment. For example, the difference amplitude of the second continuous segment is 3, 1, and 1 respectively. When it changes from 1 to 2, it is identified as an interruption point. When it is opposite to the overall downward trend of the segment, the delay value 2 can be obtained. After the above actions, all reversal delay descriptions of each pixel humidity sequence are recorded segment by segment to obtain the humidity change delay trend value.

[0079] S303: Based on the change segments in the humidity change delay trend value where the trend has not returned to the position, filter the cell areas with consecutive numbers, synchronously mark the time coordinates of the corresponding blocks, add jump trend direction information, and obtain the humidity response jump area annotation layer.

[0080] First, the humidity change trend data of each pixel in the humidity cycle sequence group is retrieved. In each humidity sequence group, a reversal signal is detected after the growth phase; that is, it is determined whether the maximum value in the time series appears in the middle or later segment. If no decline segment appears, it is considered a region where the trend has not reversed. Then, based on the trend value results, the corresponding pixel number is extracted and marked in a two-dimensional spatial raster. Simultaneously, the time index value of the first occurrence of the abnormal trend in the time series is recorded, forming an initial set of changed pixel labels with time-series information. Next, this initial label set is sorted in row and column order, and adjacent pixel combinations with consecutive numbers and a row-to-column difference of no more than 1 are defined. During the combination process, if a certain consecutive... If a break occurs in the numbering segment, areas with a numbering interval greater than 2 are discarded. In the end, sub-regions with consecutive numbers and close spatial connections are retained. In the example, if a region is numbered consecutively from 112 to 120, its spatial position row and column difference must be continuously progressive within the allowable range. There cannot be a jump in numbering such as 112, 113, and 118. Subsequently, humidity sequence trend direction indicators are added to each consecutive sub-region. The trend direction is determined based on the continuous change slope of the humidity change segment. If the difference between three consecutive periods is positive, it is an increase; otherwise, it is a decrease or fluctuation. Finally, a layer label set with pixel number, time index, and trend direction as triplets is constructed to obtain the humidity response jump area label layer.

[0081] The specific steps of S4 are as follows:

[0082] S401: Based on the location of the region in the humidity response jump area labeling layer, read the NDVI value of the corresponding pixel from the remote sensing layer periodically, arrange the periodic records in chronological order, divide the continuous NDVI amplitude sequence by region, and obtain the NDVI time series amplitude group.

[0083] First, the spatial coordinates and time markers of each pixel recorded in the layer are retrieved. Each spatial location is indexed and mapped sequentially to the image data of different periods in the remote sensing layer. The NDVI values ​​of the corresponding pixels are read sequentially, and sorted by reading time to generate a time-series dataset for each pixel. For example, for a pixel numbered 233, its NDVI values ​​for different periods form a sequence trajectory of [0.32, 0.41, 0.38]. Then, the above operation is repeated for all pixels marked in the annotation layer to construct a time-series NDVI set for all pixels. Finally, each sequence is... The pixels are aggregated and divided into subsets of regions according to their spatial numbers. For the pixel sequence within the same region, the pixels are sorted in ascending order of their numbers and recorded as the NDVI time series segments within the region. During the sorting, it is ensured that the time labels are continuous, the spatial positions are adjacent, and there are no missing values ​​or jumps. For example, if a region contains pixels numbered 201 to 208 and the time axis consists of multiple observation points, then each pixel must completely match the NDVI value of the corresponding period. Missing or empty pixels will be removed from the sequence. Finally, based on the numbering of all regions, the NDVI values ​​of each group are aggregated according to the spatial regions to obtain the NDVI time series amplitude group.

[0084] S402: Call the sequence content in the NDVI time series amplitude group, identify the changing segments with continuous growth characteristics, connect the position chains with the same direction of change, mark the start and end index positions between continuous segments, and obtain the NDVI growth trend path segments.

[0085] First, based on the NDVI time-series data corresponding to each region, starting from the initial time point, the NDVI values ​​are progressively checked to see if there is an increasing relationship. During this process, the NDVI value at the current time point is compared with the value at the previous time point. If the current value is greater than or equal to the value at the previous time point, it is determined to be in a continuous growth phase. For example, if the NDVI values ​​of pixels in a certain region are [0.27, 0.31, 0.34, 0.30, 0.36], then the first growth interval can be identified as 0.27 to 0.34, and the second growth interval after the break is 0.30 to 0.36. The starting and ending indices of this sequence are marked as [0, 2] and [3, 4], respectively. The starting and ending indices of all continuous growth segments are recorded sequentially. The endpoint is then determined. Subsequently, in the spatial dimension, the growth segments are mapped back to their respective spatial locations based on the cell number. It is then identified whether adjacent cells belong to the same time period growth trend segment. If multiple cells have overlapping time growth segments and are spatially adjacent, they are connected into a group of path segments. For example, if the three groups of cells numbered 102, 103, and 104 have NDVI segment growth times from period 1 to 3 and are three consecutive rows of cells, they are connected into a group of growth path segments. This judgment logic is repeated to mark all regions with the same continuous growth direction, consecutive numbers, and spatial connections as complete path segments. Finally, the index position information of the start and end points of each path segment in the time series is extracted to obtain the NDVI growth trend path segments.

[0086] S403: Based on the temporal order of NDVI growth trend path segments, the continuous path segments are connected in sequence, and the path structure in the growth direction is extended to adjacent areas. The connecting positions and sequence numbers are connected to obtain the vegetation main response path sequence group.

[0087] First, determine the start and end time indices and spatial locations of each path segment. Read the starting cell number of each path segment and its corresponding position index in the time series, and establish a sequence queue sorted in ascending order by the starting time. Then, perform a pairing operation on adjacent path segments sequentially, determining whether the subsequent path segment is temporally adjacent to the end time of the current segment and spatially close to the cell number. If the interval between the starting number of the path segment and the ending number of the previous segment does not exceed one cell row and column, and the difference between the start and end indices in the time series is 1, then the two are considered to be connected as a continuous path. For example, if the current path segment number is [201, 202, 203], and the time period is [4, 5, 6], and the subsequent segment number is [204, 205], and the time period is [7, 8], then the two satisfy the connection between the number and time. Once the conditions are met, a connection can be established in the path structure. The system then checks whether the starting point of the next path segment and the current ending point meet the connection conditions. If there is a gap in the path segment or the number jump exceeds the preset range (e.g., the number difference is greater than 2 or the time index is broken for more than 1 period), the connection will not be established and a new path chain will be restarted. During the path connection process, the current starting number, ending number, starting time, and ending time need to be recorded. In each connection operation, the number of the connected pixels is added to the path queue. At the same time, each path segment is assigned a unique sequence number label, such as path chain 001, path chain 002. The main response direction is marked according to the number order within the path. Finally, all path segments with spatial continuity and temporal sequence connection are connected in series, and their pixel number list and the time series range they cross are labeled to obtain the vegetation main response path sequence group.

[0088] The specific steps of S5 are as follows:

[0089] S501: Based on the path edge position in the vegetation main response path sequence group, the corresponding coordinates of the path endpoints and turning points in the remote sensing layer are calibrated, and the coordinates are matched with the node positions on the original boundary line to obtain the path boundary mapping corresponding set.

[0090] First, the starting point, ending point, and turning points at direction changes are extracted from each path. The coordinate data corresponding to these feature points are retrieved from the remote sensing layer by locating the pixel numbers of each point along the path and converting them to geographic coordinates based on the spatial reference information of the remote sensing layer. A path point coordinate table is formed by identifying the path number, row and column position, and time index of each feature point. During the path point extraction process, a coordinate set corresponding to each path is constructed using continuous segments of the vegetation response path as units, further distinguishing between start and end edge points and mid-point turning points. Then, the coordinates of each node in the original boundary line are extracted, and the coordinates of the path feature points are compared with the boundary nodes one by one. By comparing the coordinate differences point by point, it is determined whether... If a path point falls within the set matching threshold range and is deemed a match, the path point and boundary node are indexed and bound. To avoid false matches during the comparison process, if multiple path points may be close to the same boundary node, the point with the earlier time position in the path is selected for recording. If there is a dense distribution of turning points in the path, the path numbering order is used to restrict jump matching, thereby improving the accuracy of the correspondence. After completing the boundary position matching of all path feature points, the feature nodes in each path segment are stored according to the path sequence number and their corresponding boundary node indexes. Finally, a set of correspondence tables is compiled to record the spatial projection relationship of each boundary segment in the path sequence on the original boundary line of the remote sensing layer, thus obtaining the path boundary mapping correspondence set.

[0091] S502: Based on the node order in the path boundary mapping corresponding set, compare the position sequence of the path line with the original boundary line, replace the position content of the corresponding line segment on the original boundary line with the path node, synchronize the position order of the path line with the original boundary arrangement, and obtain the boundary path replacement sequence.

[0092] First, the correspondence between each path node and the original boundary segment is extracted from the mapping table. Each point in the path sequence is sorted according to its index order within the path. Using the established path-to-boundary node index mapping, the starting and ending line segment coordinates of the corresponding path node in the original boundary line are sequentially found. During the replacement process, the content of the original boundary segment is replaced according to the spatial sequence of the path line itself. All intermediate nodes between the starting and ending nodes on the original boundary line are removed, retaining only the continuous node coordinate set extracted from the path to form the replacement segment. Before replacement, it is necessary to determine whether there are discontinuous nodes in the replacement segment of the original boundary line. If there are node pairs with a jump distance exceeding a set spatial threshold, the replacement process for that segment should be stopped, and the original boundary segment should be retained. The spatial threshold can be set to a distance of less than 5 pixels between two adjacent nodes. In actual operation... If the node coordinates in the path are (102, 304), (103, 305), and (104, 306) respectively, and the corresponding nodes in the original boundary line are (102, 304) to (105, 307), then during the replacement, delete the intermediate nodes (103, 304) and (104, 305) in the original boundary segment, retain the coordinates of the path segment and write them to this position. After the replacement, the content of this segment in the original boundary line is updated to the coordinate order of the path segment. Replace each segment in sequence according to the path number on the entire boundary line. If a path segment crosses multiple boundary blocks, it needs to be split and processed to correspond to the boundary segments of their respective regions. At the same time, ensure that the segments are connected continuously after the replacement, without any coordinate jumps or direction reversals. After all path segments have been replaced, renumber the entire boundary line segment and generate a node sequence arranged in chronological order to finally obtain the boundary path replacement sequence.

[0093] S503: Based on the line segment order in the boundary path replacement sequence, organize each boundary segment according to the connection order of the path nodes, maintain the continuity of the layer number between the line segments, and connect each boundary path segment in sequence to obtain the remote sensing layer boundary update path set.

[0094] First, extract the start and end numbers of each alternative path segment on the original layer boundary. Arrange each node in the path into a directed boundary chain according to the recording order. Read and sort the segment number sequence in the original boundary layer, and compare the corresponding positions of each segment with the path segments. If the starting node of the path segment has the corresponding number N1 and the ending node has the corresponding number N2 in the layer, then renumber all segments between N1 and N2 in the original layer with the sequential node numbers in the new path segment. At the same time, replace the layer node content at these positions in the original layer with the coordinate point values ​​in the new path. If there are jump points in the path segment with a node spacing greater than two pixel widths, the consecutive numbers should be retained during recording, but the jump parts should be interpolated to ensure that the node numbers are continuous and do not have null values. During this process, each After processing a path replacement segment, the update task of a boundary path segment is completed. After all replacement path segments are processed, the entire boundary chain is reorganized and rearranged in ascending order of path number. Each segment is checked to see if the end point of each segment is connected to the start point of the next segment. If there is a breakpoint, the distance is judged by the difference between the coordinates of the end point and the start point. If the difference is less than one pixel width, it is directly connected. If it exceeds two pixel widths, the intermediate coordinate points are added by re-interpolation. When connecting, the map sheet index of the new boundary segment is recorded according to the original layer map sheet number to ensure the continuity of the numbering and the complete connection of the spatial position of the connection result. The task of re-splicing each segment on the layer boundary is completed, and after being reordered according to the spatial direction encoding method, it is recorded in the updated boundary layer file, and finally the remote sensing layer boundary update path set is obtained.

[0095] Please see Figure 2 This invention also provides an integrated space-air-ground remote sensing monitoring system for eco-hydrological elements, comprising:

[0096] The slope aspect analysis module acquires the surface slope aspect data layer of the ecological remote sensing area, selects the dominant direction of surface runoff as the reference line, compares the slope direction of adjacent pixels along the reference direction, identifies continuous jump positions, and expands to the upstream and downstream directions in combination with the spatial connection relationship of pixels. Gradually, it collects pixels with abrupt slope changes and connection nodes to obtain candidate area fragments of fracture boundaries.

[0097] The fracture extraction module extracts internal boundary line segments based on fracture boundary candidate region fragments, obtains direction vectors, arranges the direction change sequence, groups continuous vectors according to the change order, and filters boundary path segments with consistent direction change frequency and continuous distribution to obtain a set of stable boundary path segments.

[0098] The humidity response module, based on a set of stable boundary path segments, extracts the soil moisture index within the periodic path region, locates the humidity change trajectory of the pixel time series, identifies segments with continuous and consistent directions, retrieves the location of stagnant change, locates the area block whose trend has not returned to the initial state, and obtains the humidity response jump area annotation layer.

[0099] The vegetation path module extracts the NDVI amplitude change information of the region periodically based on the humidity response jump area annotation layer, filters the continuously rising path segments according to the pixel order, connects adjacent rising paths and extends along the adjacent direction, and connects the path segments in time order to obtain the vegetation main response path sequence group.

[0100] The boundary reconstruction module, based on the vegetation master response path sequence group, maps the path edge positions to the remote sensing layer boundary lines, maps the path lines to the boundary lines in sequence, replaces the boundary line segments one by one, and connects the boundary segments along the path nodes in sequence to obtain the remote sensing layer boundary update path set.

[0101] For ease of explanation, Figure 2 Only the main components of the system are shown. The system of this embodiment can be used to perform... Figure 1 The technical solutions of the method embodiments shown are similar in principle and in effect, and will not be described again here.

[0102] In this invention, slope abrupt changes are extracted by guiding the dominant runoff direction, enhancing the spatial perception of the location of terrain boundary changes. Paths are grouped according to the trend of directional vector changes to improve the continuity and stability of boundary extraction. Time labeling is performed by combining trend offset blocks in the humidity cycle to strengthen the temporal comparison in the humidity response process. Vegetation response chains are constructed by continuously splicing the NDVI growth path. Path mapping promotes the dynamic adjustment of the remote sensing boundary structure, making the boundary update process more in line with the spatiotemporal progression characteristics of ecological disturbance.

[0103] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A space-air-ground integrated remote sensing monitoring method for eco-hydrological elements, characterized in that, Includes the following steps: S1: Obtain the surface slope aspect layer, select the dominant runoff direction as the reference line, compare the slope values ​​of adjacent pixels along the direction, extract continuous jump points, extend the upstream and downstream range, record the abrupt change location, and obtain the candidate region fragment of the fault boundary. S2: Based on the candidate region segments of the fracture boundary, extract the line segment direction vectors, sort the direction change values, group them according to the change trend, filter the continuous boundary lines with consistent direction frequency repetition, retain the corresponding pixels, and obtain a set of stable boundary path segments. S3: Based on the set of stable boundary path segments, extract the soil moisture periodic sequence within the region, locate records of unrecovered trends, and combine the continuous stagnant blocks marked on the time axis to obtain a labeling layer of moisture response jump areas; S4: Based on the labeled layer of the humidity response jump area, extract the NDVI amplitude rise segment periodically, track the NDVI growth path in the adjacent direction, arrange the path segments according to time, and obtain the vegetation main response path sequence group. S5: Based on the vegetation main response path sequence group, map the path edge to the remote sensing layer boundary, replace the original line segment according to the path and reconnect the cell boundary, and connect them to the whole map in sequence to obtain the remote sensing layer boundary update path set.

2. The integrated space-air-ground remote sensing monitoring method for eco-hydrological elements according to claim 1, characterized in that, The candidate region segment of the fracture boundary includes pixels with abrupt changes in slope direction, groups of nodes with continuous jumps, and upstream and downstream connected pixel blocks. The set of stable boundary path segments includes a sequence of direction vector arrangements, hierarchical grouping of direction changes, and continuously distributed boundary path segments. The labeling layer of the humidity response jump region includes soil humidity time trajectory, humidity trend turning point, trend offset patch area, and time series label. The vegetation main response path sequence group includes NDVI amplitude rising segment, continuously growing path chain, neighborhood extension path segment, and time series connection channel line. The set of remote sensing layer boundary update paths includes boundary mapping reference point, path position replacement node, sequential reconnection boundary chain segment, and layer boundary connection trajectory.

3. The integrated space-air-ground remote sensing monitoring method for eco-hydrological elements according to claim 1, characterized in that, The continuous boundary line of the path refers to the pixel path segments that have the same trend of directional change, the same frequency of direction, and are arranged continuously in space in the candidate region of the fracture boundary. The continuous stagnant block refers to a pixel region segment in the humidity sequence where the direction of change has not been restored and the region remains stagnant and without reversal on the time axis.

4. The integrated space-air-ground remote sensing monitoring method for eco-hydrological elements according to claim 1, characterized in that, The term "trend not recovered record" refers to the time point in the humidity cycle sequence where a pixel does not return to the original trend direction after experiencing a trend change. The pixel boundary refers to the spatial boundary line formed by connecting adjacent pixels in the remote sensing layer.

5. The integrated space-air-ground remote sensing monitoring method for eco-hydrological elements according to claim 1, characterized in that, The specific steps of S1 are as follows: S101: Obtain the surface slope aspect data layer in the ecological remote sensing area, select the dominant direction of surface runoff as the reference line, extract the slope direction data of adjacent pixels sequentially along the reference line, compare the slope direction of the pixel with the search direction of the previous pixel, record the position sequence of the direction jump, and obtain the slope aspect jump position record group. S102: Based on the slope aspect change location record group, expand the adjacent pixel data in the three rows of the upstream and downstream of the extended region, retrieve the directional position of the continuous connection of pixels in space, mark the upstream and downstream extension zones according to the spatial continuity relationship, and obtain the slope aspect change expansion region zone. S103: Based on the slope aspect change extension zone, sequentially read the cell number and coordinates at the slope aspect change position, summarize the spatially continuous cell combinations with jump characteristics, output the number list in the recording order, and obtain the candidate region fragment of the fracture boundary.

6. The integrated space-air-ground remote sensing monitoring method for eco-hydrological elements according to claim 1, characterized in that, The specific steps of S2 are as follows: S201: Based on the boundary line segments in the candidate region of the fracture boundary, obtain the direction vector value corresponding to the line segment, connect adjacent vectors according to the line segment arrangement order, remove line segment combinations with discontinuous direction changes, and obtain a sequence of direction change values. S202: Based on the change values ​​in the numerical sequence of direction changes, calculate the rate of change of each segment of direction change difference, divide the direction change interval, and group the paths whose rate values ​​fall within the same range into the same group to obtain the direction change rate interval. S203: Based on the spatial distribution of path segments in the range of directional change rates, filter out pixel segments that are adjacent in position and whose directional trend is not interrupted, exclude discontinuous path positions, and obtain a set of stable boundary path segments.

7. The integrated space-air-ground remote sensing monitoring method for eco-hydrological elements according to claim 1, characterized in that, The specific steps for S3 are as follows: S301: Based on the region corresponding to the set of stable boundary path segments, collect the soil moisture index of the pixels within the period, arrange the moisture data at the same position in chronological order, and distinguish the storage sequence trajectory according to the pixel number to obtain the soil moisture periodic sequence group. S302: Based on the humidity sequence corresponding to the pixel in the soil moisture periodic sequence group, identify the segments in the sequence with continuous and consistent change direction, find the fluctuation interruption point in the continuous segment and record the position change trend, calculate the reversal delay value between the change amplitude and the trend direction in each segment of the sequence, and obtain the humidity change delay trend value. S303: Based on the fact that there are change segments in the humidity change delay trend value where the trend has not returned, filter out the pixel areas with consecutive numbers, synchronously mark the time coordinates of the corresponding blocks, add jump trend direction information, and obtain the humidity response jump area annotation layer.

8. The integrated space-air-ground remote sensing monitoring method for eco-hydrological elements according to claim 1, characterized in that, The specific steps of S4 are as follows: S401: Based on the location of the region in the humidity response jump region labeling layer, read the NDVI value of the corresponding pixel from the remote sensing layer periodically, arrange the periodic records in chronological order, divide the continuous NDVI amplitude sequence by region, and obtain the NDVI time series amplitude group. S402: Call the sequence content in the NDVI time series amplitude group, identify the changing segments with continuous growth characteristics, connect the position chains with the same direction of change, mark the start and end index positions between continuous segments, and obtain the NDVI growth trend path segments. S403: Based on the time sequence of the NDVI growth trend path segments, the continuous path segments are connected in sequence, and the path structure in the growth direction is extended to the adjacent area. The connecting positions and sequence numbers are connected in series to obtain the vegetation main response path sequence group.

9. The integrated space-air-ground remote sensing monitoring method for eco-hydrological elements according to claim 1, characterized in that, The specific steps of S5 are as follows: S501: Based on the path edge position in the vegetation main response path sequence group, mark the corresponding coordinates of the path endpoints and turning points in the remote sensing layer, match the coordinates with the node positions on the original boundary line, and obtain the path boundary mapping corresponding set. S502: Based on the node order in the path boundary mapping corresponding set, compare the position sequence of the path line with the original boundary line, replace the position content of the corresponding line segment on the original boundary line with the path node, synchronize the position order of the path line with the original boundary arrangement, and obtain the boundary path replacement sequence. S503: Based on the line segment order in the boundary path replacement sequence, organize each boundary segment according to the connection order of the path nodes, maintain the continuity of the layer number between the line segments, and connect each boundary path segment in sequence to obtain the remote sensing layer boundary update path set.

10. A space-air-ground integrated remote sensing monitoring system for eco-hydrological elements, characterized in that, The system is used to implement the integrated space-air-ground remote sensing monitoring method for eco-hydrological elements as described in any one of claims 1-9, and the system includes: The slope aspect analysis module acquires the surface slope aspect data layer of the ecological remote sensing area, selects the dominant direction of surface runoff as the reference line, compares the slope direction of adjacent pixels along the reference direction, identifies continuous jump positions, and expands to the upstream and downstream directions in combination with the spatial connection relationship of pixels. Gradually, it collects pixels with abrupt slope changes and connection nodes to obtain candidate area fragments of fracture boundaries. The fracture extraction module extracts internal boundary line segments based on the candidate fracture boundary region fragments, obtains direction vectors, arranges the direction change sequence, groups continuous vectors according to the change order, and filters boundary path segments with consistent direction change frequency and continuous distribution to obtain a set of stable boundary path segments. The humidity response module, based on the set of stable boundary path segments, extracts the soil moisture index within the path area period, locates the humidity change trajectory of the pixel time series, identifies segments with continuous and consistent directions, retrieves the position of stagnant change, locates the area block where the trend has not returned to the initial state, and obtains the humidity response jump area annotation layer. The vegetation path module extracts the NDVI amplitude change information of the region periodically based on the humidity response jump area annotation layer, filters the continuously rising path segments according to the pixel order, connects adjacent rising paths and extends along the adjacent direction, and connects the path segments in time order to obtain the vegetation main response path sequence group. The boundary reconstruction module maps the path edge positions to the remote sensing layer boundary lines based on the vegetation main response path sequence group, matches the path lines with the boundary lines in sequence, replaces the boundary line segments one by one, and connects the boundary segments along the path nodes in sequence to obtain the remote sensing layer boundary update path set.

Images