Ecological remote sensing data processing system

Abstract

The invention relates to the technical field of geographic information data, in particular to an ecological remote sensing data processing system which comprises a remote sensing layer acquisition module, a boundary position checking module, a region coding integration module, a spatial change splicing module and an ecological trend drawing module. According to the method, by accurately matching the geographic identification information and the spatial boundary value, the coverage range and the positioning precision of the ground feature pattern spots are remarkably improved, the time information and the spatial information are effectively integrated, the continuity and the accuracy of the data are enhanced, the boundary resetting technology effectively reduces the errors of the geographic information, and the time-space continuity of the data is improved; the processing efficiency is greatly improved by constructing a continuous layer structure and optimizing a data retrieval process, the fine boundary splicing technology also allows fine tracking of ecological changes, more detailed data support is provided for ecological monitoring, the ability of ecological trend analysis is enhanced, and thus the overall effect and application value of ecological monitoring are improved.

Description

Technical Field

[0001] The present invention relates to the field of geographic information data technology, and in particular to an ecological remote sensing data processing system. Background Art

[0002] The field of geographic information data technology encompasses the collection, storage, processing, and analysis of spatial location information and related attributes. This technology integrates remote sensing imagery, environmental monitoring data, and ecological elements based on geographic coordinate systems to construct a multidimensional spatiotemporal database. Its core focus is supporting natural resource assessment and ecological change tracking through coordinate registration, data fusion, and spatial modeling. This involves satellite data interpretation, ground observation data correction, and the calculation of multi-scale ecological indicators. It serves applications such as national land planning, disaster warning, and biodiversity conservation. Current technology is limited by low heterogeneous data integration efficiency, insufficient spatiotemporal continuity, and discrepancies in data standardization.

[0003] Among them, the ecological remote sensing data processing system refers to a technical system for the coordinated processing of remote sensing data and environmental parameters in ecological monitoring scenarios. By establishing an automatic matching mechanism between multi-source remote sensing data and ground observation data, a unified spatiotemporal benchmark is formed, a dynamic monitoring framework for ecological elements is constructed, and data partitioning management technology based on geographic grids is used to achieve rapid indexing of massive data. Specifically, coordinate conversion rules are used to eliminate spatial deviations caused by sensor differences, and a data hierarchical storage structure is used to separate basic geographic information and ecological thematic attributes. Based on spatial overlay analysis methods, the spatial distribution characteristics of ecological indicators such as vegetation cover and soil erosion are extracted. Based on standardized data formats, time series comparison of multiple periods of data is achieved, and spatial interpolation technology is used to supplement the ecological parameters of missing areas. Ecological assessment results of different scales are output through thematic mapping templates.

[0004] Existing technologies are inefficient when processing large-scale heterogeneous data. The lack of an efficient automatic matching mechanism and a unified spatiotemporal benchmark often leads to inaccurate spatial correspondence and inconsistent levels during data integration. Spatial deviations between remote sensing data and ground observation data can cause errors in the assessment of vegetation cover and soil erosion, thereby affecting the accuracy of decision-making. Differences in data standardization and spatiotemporal discontinuities can also lead to breakpoints in the monitoring of ecological change trends, making the analysis of long-term ecological changes incomplete, limiting the effective assessment of ecosystem health, and thus affecting the formulation of ecological protection and resource management strategies. Summary of the Invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose an ecological remote sensing data processing system.

[0006] In order to achieve the above-mentioned purpose, the present invention adopts the following technical solution: an ecological remote sensing data processing system includes:

[0007] The remote sensing layer acquisition module obtains the images scanned by the remote sensing platform, combines the image geographic information with the boundary value, identifies the covered objects, matches the known identifiers, constructs the map area, marks the upper, lower, left and right boundary points, and generates a bounding box information set;

[0008] The boundary position verification module extracts the projection plane position based on the bounding box information set, calls the map reference boundary marking, performs boundary reset according to the offset direction and distance of the diagonal points of the map spot and the reference boundary, relocates the four corners of the map spot, and generates a relocated map spot coding layer;

[0009] The regional coding integration module calls the relocated patch coding layer, constructs the layer in chronological order, retrieves patches with continuous and adjacent coding, packages them into spatial segments, and generates a combined boundary structure set;

[0010] The spatial variation stitching module calls the combined boundary structure set, selects vegetation and wetland blocks, compares the overlap of adjacent block outlines, determines boundary drift, records outline numbers, and generates a stitching block sequence;

[0011] The ecological trend drawing module calls the mosaic tile sequence, retrieves repeated extension directions and expansion forms, matches degradation zones and variation points, combines convergent tiles into trend segments, and obtains an ecological remote sensing dataset.

[0012] As a further solution of the present invention, the boundary box information set includes the upper left boundary coordinates, lower right boundary coordinates, lower left boundary coordinates, and upper right boundary coordinates of the feature map; the repositioned map coding layer includes a projection boundary plane coordinate set, a reference boundary marking space parameter, an offset direction identifier, and a distance threshold parameter; the combined boundary structure set includes a geocoding unique identifier, a timestamp index, an adjacent map spatial topological relationship, and a fragment combination closed boundary line; the spliced ​​block sequence includes a drift judgment result table, a contour offset matrix, and a junction area splicing path vector set; the ecological remote sensing data set includes the spatial trajectory coordinates of the degraded contiguous belt, the variation concentration area polygon vertex set, the trend fragment vector overlay unit, and the extension direction repeatability identifier.

[0013] As a further solution of the present invention, the remote sensing layer acquisition module includes:

[0014] The image geolocation submodule collects ground scanning image data from the remote sensing platform, extracts the geographic identification information carried by each image, compares the corresponding positions of the coordinate points, image time stamps and spatial boundary values ​​in the image, calculates the spatial mapping point distribution of the image data, obtains the location of the image in geographic space, and generates the image spatial positioning coefficient value;

[0015] The patch boundary recognition submodule calls the area range within the boundary value in the image based on the image spatial positioning coefficient value, calculates the similarity coefficient between the pixel characteristic value and spatial distribution value in the area and the spatial identification of the known ground object, and judges and screens the patches in the area according to the similarity coefficient and the set spatial matching threshold, obtains the spatial partition value of the area covered by the boundary range, and generates a spatial patch matching value;

[0016] The bounding box generation submodule identifies the contour edge of the covered pattern according to the spatial pattern matching value, determines the upper, lower, left, and right boundary point positions of the edge segment in the image pixel grid, calculates the coordinate values ​​of the boundary point set, and uses the formula:

[0017]

[0018] Obtain boundary envelope trend strength values ​​through calculation, construct boundary point sets based on the boundary envelope trend strength values, obtain closed graphic areas of the patch boundaries, and generate boundary box information sets;

[0019] Among them, B ij Represents the boundary envelope trend strength value, C ik Represents the boundary coordinate difference of the kth point in the i-th area of ​​the edge of the spot contour, D jk Represents the center extension distance of the kth point in the jth area in the edge of the pattern contour, M ijk Represents the weighted matching factor between the boundary point and the matching value, P i Represents the horizontal coordinate difference between the upper boundary point and the center, Q j Represents the vertical coordinate difference between the lower boundary point and the center, and n represents the number of boundary points involved in the calculation in each area.

[0020] As a further solution of the present invention, the boundary position verification module includes:

[0021] The boundary point extraction submodule extracts the projection plane boundary position used by the patch based on the boundary points marked in the bounding box information set, identifies the coordinate data and spatial position index corresponding to the boundary points, performs boundary point integrity detection and patch boundary topological relationship judgment, and obtains the boundary coordinate index value;

[0022] The diagonal offset calculation submodule calls the boundary coordinate index value, selects the diagonal points of the patch corresponding to the reference boundary according to the set of patch boundary points corresponding to the boundary coordinate index value, collects the horizontal and vertical coordinate offset values, analyzes the spatial position change trend between the diagonal points of the patch and the reference mark, and obtains the patch offset angle value set;

[0023] The boundary position resetting submodule adjusts the four position coordinates of the boundary of the patch according to the offset direction and offset distance of the patch in the patch offset angle value set, performs coordinate difference correction and spatial reprojection operations, and generates a relocated patch coding layer.

[0024] As a further solution of the present invention, the region code integration module includes:

[0025] The patch coding acquisition submodule calls the relocated patch coding layer to obtain the geographic code and acquisition time stamp of each patch, extracts its latitude and longitude coordinate values ​​and timestamp values ​​according to the unique identification value of the patch, and uses the patch ID as the index field to structure the collected parameters, construct a patch parameter set table, and generate a patch coding information set;

[0026] The coding sequence construction submodule calls the patch coding information set, sorts all patches according to their acquisition time stamps, reconstructs the patch data sequence in chronological order, identifies the coding differences and spatial adjacency between adjacent patches, groups patches with continuous coding and spatial adjacency, and generates a set of adjacent patch sequences.

[0027] The spatial fragment combination submodule extracts the boundary coordinate points of each group of patches based on the adjacent patch sequence set, calculates the boundary overlap and cross-extension value of each combined patch, and determines the integrity boundary value of its fragment combination using the formula:

[0028]

[0029] The calculation obtains the combined boundary consistency value, and the corresponding patch combination is selected according to the result that the boundary consistency value is greater than the combined boundary threshold, and its closed boundary path is established to generate the combined boundary structure set;

[0030] Among them, B s represents the combined boundary consistency value, C z Represents the number of boundary points of the zth patch, L z Represents the length of the boundary line segment of the zth spot, A z Represents the area value of the zth spot, M z Represents the boundary overlap distance between the zth spot and its adjacent spots. Represents the average spatial adjacent distance value of the entire patch group, D u represents the outer boundary extension length of the u-th group of patches, T u Represents the number of patches contained in the u-th group of patches, V u represents the internal boundary jump degree of the u-th group of patches, w represents the total number of patches contained in the currently processed patch combination, and m represents the number of external boundary extension groups involved in the current patch combination.

[0031] As a further solution of the present invention, the space-varying splicing module includes:

[0032] The boundary line extraction submodule calls the combined boundary structure set, obtains the original layer content of the vegetation coverage area tile and the wetland distribution area tile, detects the continuous time layer frames of the tile, calls the boundary line data of the tile in the adjacent time frames, compares the boundary direction differences of the tile in the adjacent time frames, numbers the outline of each frame tile, and generates a contour boundary number sequence;

[0033] The contour overlap determination submodule compares the geometric overlap rates of the boundary lines of adjacent tiles in the continuous layer according to the contour boundary number sequence, calculates the overlap rate value and spatial offset distance value of each pair of adjacent tile contour boundaries, and determines whether the set contour offset threshold and spatial position offset threshold are exceeded using the formula:

[0034]

[0035] The operation obtains the spatial offset value of the block. Based on the situation where the offset threshold is exceeded, each block with boundary drift behavior is obtained, and the starting contour number and the ending contour number are recorded to obtain a list of boundary drift contour intervals.

[0036] Among them, R disp Represents the tile space offset value, L e represents the length of the boundary line of the e-th block, ΔC e Represents the displacement of the contour center between the two frames before and after the e-th block, O e Represents the boundary overlap value of the e-th block, S e represents the average offset of the boundary line of the e-th tile, and g represents the total number of tile contour numbers;

[0037] The mosaic tile generation submodule calls the boundary drift contour interval list, screens contour combinations whose spatial overlap is greater than the tile mosaic reference threshold based on the boundary line continuity between contours and the tile boundary coordinates, sorts the contour numbers to form the tile connection order, obtains the tile boundary correspondence, and generates a mosaic tile sequence.

[0038] As a further solution of the present invention, the ecological trend drawing module includes:

[0039] The boundary splicing submodule calls the boundary information of each tile in the splicing tile sequence, extracts the spatial contour edge segments, compares the splicing relationship of the edge segments of adjacent tiles, constructs a splicing path, and collects the extension trend of the path in two-dimensional space to generate a tile splicing extension line;

[0040] The direction recognition submodule calls the block splicing extension line, extracts the extension path set of multiple time periods, calculates the main axis vector of the path on the spatial coordinate axis, compares the main axis vector directions of different time periods within the same block, and selects the direction distribution lines with repeated directions reaching a set number of times to obtain the repeated extension direction value;

[0041] The trend combination submodule selects line segments with consistent directions in the extension line according to the repeated extension direction values, extracts the spatial block identifiers corresponding to the extension directions, determines the spatial overlap between the distribution lines of the degradation contiguous belts and the positioning points of the variation concentration areas in the ecological scene, screens the line segment combinations with overlaps exceeding the set threshold, splices them into continuous spatial segments, and generates the ecological remote sensing dataset.

[0042] Compared with the prior art, the advantages and positive effects of the present invention are:

[0043] In the present invention, by accurately matching geographic identification information and spatial boundary values, the coverage and positioning accuracy of the ground feature map are significantly improved, the time and space information are effectively integrated, and the continuity and accuracy of the data are enhanced. The boundary reset technology effectively reduces the error of geographic information and improves the spatiotemporal continuity of the data. The construction of a continuous layer structure and the optimization of the data retrieval process greatly improve the processing efficiency. The fine boundary splicing technology also allows for detailed tracking of ecological changes, provides more detailed data support for ecological monitoring, and enhances the ability of ecological trend analysis, thereby improving the overall effect and application value of ecological monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 is a system flow chart of the present invention;

[0045] Figure 2 This is a flow chart of the remote sensing layer acquisition module of the present invention;

[0046] Figure 3 This is a flow chart of the boundary position verification module of the present invention;

[0047] Figure 4 This is a flow chart of the regional coding integration module of the present invention;

[0048] Figure 5 This is a flow chart of the spatial variation splicing module of the present invention;

[0049] Figure 6 Draw a module flow chart for the ecological trend of the present invention. DETAILED DESCRIPTION

[0050] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0051] In the description of the present invention, it should be understood that the terms "length," "width," "up," "down," "front," "back," "left," "right," "vertical," "horizontal," "top," "bottom," "inside," "outside," and the like, indicating positions or relationships, are based on the positions or relationships shown in the accompanying drawings and are intended only to facilitate the description of the present invention and simplify the description. They do not indicate or imply that the devices or elements referred to must have a specific orientation, be constructed, or operate in a specific orientation. Therefore, they should not be construed as limiting the present invention. Furthermore, in the description of the present invention, "plurality" means two or more, unless otherwise expressly and specifically defined.

[0052] See also Figure 1 , an ecological remote sensing data processing system includes:

[0053] The remote sensing layer acquisition module obtains ground scanning image data from the remote sensing platform, combines the geographic identification information and spatial boundary values ​​carried by each image, identifies the covered features in the area, and constructs the coverage range of the corresponding map patch by matching the positions with the spatial identifiers of known features. It also marks the upper, lower, left, and right boundary points of the map patch to obtain the bounding box information set.

[0054] The boundary position verification module extracts the boundary position of the projection plane used based on the boundary points marked in the bounding box information set, calls the reference boundary markings in the target map, performs boundary reset operations according to the offset direction and distance between the diagonal points of the map patch and the reference boundary, locates the new four corner positions of each map patch, and obtains the relocated map patch coding layer;

[0055] The regional coding integration module calls the geographic code and acquisition time stamp of each patch in the relocated patch coding layer, builds a continuous layer structure in the order of shooting time, retrieves patches with continuous codes and spatial adjacency, and packages them into spatial fragment combinations. For each combination, a spatial boundary line is established to obtain a combination boundary structure set.

[0056] The spatial variation stitching module combines the continuous temporal content and corresponding boundary lines of the fragments in the combined boundary structure set, selects the vegetation coverage area tiles and wetland distribution area tiles in the ecological remote sensing scene as the combination objects, compares the degree of overlap of the contour positions of adjacent tiles in the continuous layer, determines whether there is boundary drift, records the start and end contour numbers of the tiles that have drifted, and generates a stitching tile sequence based on the tile boundary area;

[0057] The ecological trend drawing module calls the splicing paths and spatial extension lines of the tiles in the splicing tile sequence, retrieves the repeated extension directions and regional expansion forms that appear in multiple time periods, matches the distribution lines of degradation belts and the positioning points of variation concentration areas in the ecological scene, and combines the spatial segments with repeated extensions and converging boundaries in the tiles into trend segments to obtain the ecological remote sensing dataset.

[0058] The bounding box information set includes the upper left boundary coordinates, lower right boundary coordinates, lower left boundary coordinates, and upper right boundary coordinates of the feature patch. The relocated patch coding layer includes the projection boundary plane coordinate set, reference boundary marking spatial parameters, offset direction identifier, and distance threshold parameters. The combined boundary structure set includes the geocoding unique identifier, timestamp index, spatial topological relationship of adjacent patches, and fragment combination closed boundary line. The spliced ​​tile sequence includes the drift judgment result table, contour offset matrix, and junction area splicing path vector set. The ecological remote sensing data set includes the spatial trajectory coordinates of the degraded contiguous belt, the polygon vertex set of the variation concentration area, the trend fragment vector overlay unit, and the extension direction repeatability identifier.

[0059] See also Figure 2 , the remote sensing layer acquisition module includes:

[0060] The image geolocation submodule collects ground scanning image data from the remote sensing platform, extracts the geographic identification information carried by each image, compares the corresponding positions of the coordinate points, image time stamps and spatial boundary values ​​in the image, calculates the spatial mapping point distribution of the image data, obtains the relative positioning results of the image in geographic space, and generates the image spatial positioning coefficient value;

[0061] When collecting ground scanning image data equipped with a remote sensing platform, it is necessary to specify the scanning path range and the corresponding scanning resolution parameters. For example, the platform is set to scan in the north-south direction with a spacing of 10 meters and a spatial resolution of 0.5 meters. The acquisition time of each image is recorded according to the image acquisition timestamp, and the original GPS coordinate information recorded by the image sensor is bound at the same time. For example, the latitude and longitude coordinates of the image collected at a certain moment are (117.1524°E, 39.1236°N). Through coordinate correction, it is projected into the preset UTM projection system, and the mapping coordinates become (457200.3, 4336200.1). The mapping matrix is ​​established by corresponding it to the row and column numbers of the image pixels. The image boundary values ​​are then parsed based on this matrix. For example, if the image width and height are 4000×3000 pixels respectively, and the projection area covers 1000×750 meters, the unit pixel mapping distance is calculated to be 0.25 meters / pixel, and the actual position of any pixel point in the geographic space is deduced. For example, the actual coordinates corresponding to the pixel in the 1500th row and 2000th column are (457200.3+500, 4336200.1+375), that is, (457700.3, 4336575.1). The above steps are performed in batches when processing image group data to traverse the geographic identification information of each pixel point in each image and finally generate the image space positioning coefficient value.

[0062] The patch boundary recognition submodule calls the area within the boundary value in the image based on the image spatial positioning coefficient value, calculates the similarity coefficient between the pixel feature value and spatial distribution value in the area and the spatial identification of the known ground object, and judges and screens the patches in the area according to the similarity coefficient and the set spatial matching threshold, obtains the spatial partition value of the area covered by the boundary range, and generates the spatial patch matching value;

[0063] Based on the image space positioning coefficient value, the specific projection position of each remote sensing image in the geographic space can be clarified, the image area within the range is called, and the pixel data in the target area is extracted. By setting a 3×3 sliding window to traverse the image grayscale value, the median, variance and range are obtained in each window as pixel feature indicators. For example, the pixel value of the center of the region is 135, and the maximum and minimum difference values ​​of the neighborhood are 20, then the range is 20. This value is compared with the standard feature value of the known spatial identification of the ground object. Assuming that the standard feature value of the known ground object type A is [130, 150], the current pixel meets the preliminary matching conditions. Further based on the average grayscale mean and spatial orientation of the regional block in the image For example, a certain patch is oriented northward, has an average grayscale of 140, and a spatial weight value of 0.9. The similarity coefficient between it and type A is calculated through the normalized spatial matching function. If the similarity is greater than the set spatial matching threshold of 0.85, it is considered a successful match. The matching threshold is set with reference to the stable value interval of spatial features in the historical area. By counting the similarity mean and standard deviation of 100 sample areas, the matching threshold is calculated as the mean minus one standard deviation, that is, threshold = 0.91-0.06 = 0.85. If the match is successful, the boundary of the patch area is delineated, and the spatial partition value corresponding to the area is obtained, and finally the spatial patch matching value is generated.

[0064] The bounding box generation submodule identifies the contour edge of the covered patch based on the spatial patch matching value, determines the upper, lower, left, and right boundary point positions of the edge segment in the image pixel grid, calculates the coordinate values ​​of the boundary point set, and uses the formula:

[0065]

[0066] Obtain boundary envelope trend strength values ​​through calculation, construct boundary point sets based on the boundary envelope trend strength values, obtain closed graphic areas of the patch boundaries, and generate boundary box information sets;

[0067] Among them, B ij Represents the boundary envelope trend strength value, C ik Represents the boundary coordinate difference of the kth point in the i-th area of ​​the edge of the spot contour, D jk Represents the center extension distance of the kth point in the jth area in the edge of the pattern contour, M ijk Represents the weighted matching factor between the boundary point and the matching value, P i Represents the horizontal coordinate difference between the upper boundary point and the center, Q j represents the vertical coordinate difference between the lower boundary point and the center, and n represents the number of boundary points involved in the calculation in each area;

[0068] According to the spatial pattern matching value, the boundary points of the pattern outline are extracted and the boundary envelope trend intensity value is calculated. After the boundary points are extracted, the horizontal and vertical coordinate differences of the upper, lower, left and right boundary points are recorded. Assuming that in area 1, the upper boundary point is (500, 650), the lower boundary point is (500, 700), the left boundary point is (470, 675), and the right boundary point is (530, 675), then the horizontal difference P i =530-470=60, longitudinal difference Q j

[0069] =700-650=50, the distance from each boundary point to the center point (500, 675) is measured in turn, and the coordinate difference C of point 1 i1 =30, extended distance D j1 =25, matching factor M ij1 =0.88, the difference at point 2 is C i2

[0070] =20, extended distance D j2 =22, matching factor M ij2 =0.86, substitute into the formula:

[0071]

[0072] Molecular calculation:

[0073] (30+25)·0.88=55·0.88=48.4;

[0074] (20+22)·0.86=42·0.86=36.12;

[0075] Numerator sum = 48.4 + 36.12 = 84.52;

[0076] Denominator calculation:

[0077]

[0078] Final calculation:

[0079]

[0080] The results show that the boundary strength value of the current patch area is 1.082, indicating that the patch outline has strong convergence and can generate a closed bounding box by corresponding to the upper, lower, left, and right boundary points. This value is within the set boundary envelope strength reference range [0.85, 1.15], which is considered valid and ultimately generates a bounding box information set. By integrating the boundary difference, center expansion, and matching factors, and adding a normalized scale denominator, the boundary strength index is made more adaptable to geometric scale and feature integration, thereby improving the accuracy of closed boundary construction.

[0081] See also Figure 3 , the boundary position verification module includes:

[0082] The boundary point extraction submodule extracts the projection plane boundary position used by the patch based on the boundary points marked in the bounding box information set, identifies the coordinate data and spatial position index corresponding to the boundary points, performs boundary point integrity detection and patch boundary topological relationship judgment, and obtains the boundary coordinate index value;

[0083] First, the coordinate data of the boundary points corresponding to each patch are extracted. This process is usually based on the patch vector file exported from the geographic information platform. Each patch has boundary points, marked by the four corner points of southwest, southeast, northeast, and northwest. The patch boundary is stored as a closed polygon. In the implementation, taking patch T001 as an example, the four corner points are (352000.25, 4070000.75), (352000.25, 4070002.65), (352002.15, 4070002. 65) and (352002.15, 4070000.75). Its boundary data is complete and meets the closure condition. When performing coordinate conversion on the patch data, a unified projection coordinate system is required. This implementation adopts the WGS84UTMZone50N coordinate system. After conversion, two decimal places are retained to ensure coordinate accuracy. The obtained coordinate accuracy range is set to ±0.05 meters. The patches outside this error range need to recalculate the original point position, and then perform boundary point integrity detection, comparing the four corner points one by one to see if there are any missing points. Or repeated records, in the patch T002, if there are two points with completely consistent coordinates or a corner point is missing, it is judged as an incomplete patch. This type of patch will not be involved in subsequent processing after being marked as an abnormal patch. In addition to the number of boundary points, it is also necessary to determine whether the lines between the points are crossed or not closed. The distance continuity judgment method is used. If the distance between two adjacent points is greater than 1.5 meters and is not closed, it is recorded as a structural abnormal patch. When judging the topological relationship, the patch and its adjacent patches need to be matched with adjacent edges. If the overlap is less than 90%, %, that is, it is judged that its boundary is not closed and continuous. The threshold value of overlap is set based on the actual distribution characteristics of the regional patches. According to the density characteristics of the patches in the urban construction area, 90% is determined to be the critical value of structural stability. This value is obtained by sampling and comparing a large number of overlapping boundary patches. For example, taking 30 groups of patches as samples, the average boundary overlap rate is 92.8% and the standard deviation is 1.3%. The lower value of 90% is selected as the judgment basis, and finally the boundary coordinate index set that completes the projection transformation and structural verification is output to obtain the boundary coordinate index value.

[0084] The diagonal offset calculation submodule calls the boundary coordinate index value, selects the diagonal points of the patch corresponding to the reference boundary according to the set of patch boundary points corresponding to the boundary coordinate index value, collects the horizontal and vertical coordinate offset values, analyzes the spatial position change trend between the diagonal points of the patch and the reference mark, and obtains the patch offset angle value set;

[0085] Call the boundary coordinate index value, and analyze the spatial relationship between each patch and the reference boundary according to the coordinates of the corner point located by the patch in the boundary coordinate index value. In actual operation, it is necessary to first lock the corresponding area range of the patch and the reference boundary to ensure the effectiveness of the comparison between patches in the same block. The southwest corner point of patch T001 (352000.25, 4070000.75) is extracted, and the corresponding reference boundary point is (352002.15, 4070002.65). The horizontal distance is 1.90 meters and the vertical distance is 1.90 meters calculated by the coordinate difference. The coordinate difference is combined with the patch direction angle. The patch direction angle θ is set to 45.0 degrees, and the reference boundary angle β is 45.2 degrees. The angle difference between the two is 0.2 degrees. According to the engineering surveying and mapping accuracy standard, the angle difference threshold is set to 0.5 degrees. That is, if the direction difference exceeds this range, it is marked as an abnormal boundary direction patch. The setting is based on the urban planning patch edge error control standard. The average angle difference in the regional sample is 0.36 degrees, so 0.5 degrees is selected as the maximum allowable deviation. The angular difference between patch T002 and the reference boundary is 0.5 degrees, which is a critical mark; the difference of T003 is 0.2 degrees, which is within the acceptable range. The corresponding offset values ​​are 2.4 meters and 1.3 meters respectively. In the offset analysis, if the patch offset exceeds 2.5 meters, it is marked as a severely offset patch. This offset distance threshold is derived from the analysis of existing patch offset historical data. The average patch offset in the urban area is 1.8 meters. The safety threshold is set to 2.5 meters as the acceptable upper limit. The values ​​of T001, T002, and T003 are 1.9, 2.4, and 1.3 respectively, which are within the set thresholds. The patch number, offset distance, and angular difference are recorded as a structural vector for subsequent position adjustment and call to obtain the patch offset angle value set.

[0086] The boundary position resetting submodule adjusts the four position coordinates of the boundary of the patch according to the offset direction and offset distance of the patch offset angle value, performs coordinate difference correction and spatial reprojection operations, and generates a relocated patch coding layer;

[0087] According to the offset deflection value of the patch, the offset direction and offset distance of each patch are concentrated, and the coordinate adjustment method is used to reposition the four positions. Before repositioning the patch, the superposition relationship between its four corner points and the offset vector must be analyzed. For example, the offset distance of the diagonal point of the T001 patch is 1.90 meters, and the direction angle is 45 degrees. According to the vector decomposition, the offset vector is decomposed into the X and Y directions, which are 1.34 meters and 1.34 meters respectively after decomposition. The X and Y coordinates of the four corner points of T001 are added, and the new boundary point coordinates are (352001.59, 4070002.09), (352001.59, 4070000.89), (352003.49, 407 0002.09), (352003.49, 4070000.89), in this process, it is necessary to ensure that the closure of the four sides remains unchanged, that is, the topological structure of the original patch is retained after the boundary is relocated, and at the same time check whether the seams between adjacent patches are broken due to relocation. If the distance between adjacent patches is less than 0.3 meters, the system automatically performs edge patch repair operations. This value is taken from the boundary accuracy error tolerance standard. The patch after boundary adjustment generates a new boundary coordinate set again through the coordinate index calculation method, and is encoded according to the patch number. At the same time, the layer attribute fields are updated, including the patch number, four corner coordinates, offset value, etc. Finally, all patches are output as a new layer file to obtain the relocated patch encoding layer.

[0088] See also Figure 4 , the regional coding integration module includes:

[0089] The patch coding acquisition submodule calls the relocated patch coding layer to obtain the geographic code and acquisition time stamp of each patch. Based on the unique identification value of the patch, its latitude and longitude coordinates and timestamp values ​​are extracted respectively. The collected parameters are structured and organized using the patch ID as the index field. A patch parameter set table is constructed to generate a patch coding information set.

[0090] First, the unique identification value of each patch is extracted one by one according to the patch coding layer. The identification value can be a number such as GID001, GID002, etc., and it is used as the index field to read the corresponding geocode and collection time stamp. The geocode consists of longitude and latitude. For example, the center point coordinates of patch GID001 are (112.3512, 24.5678), and the collection time stamp is "2023-09-1014:35:22". The longitude and latitude values ​​are directly read through the geometric attributes in the layer data structure, and the collection time is retrieved from the attribute field. Then, the above patch ID, longitude and latitude and timestamp are combined into a structured record in the form of (GID001, 112.3512, 24.5678, 20230910143522). This process traverses and extracts all patches in the layer. When the layer contains 500 patches, it is necessary to loop 500 times to capture parameters and fill them into the patch parameter set array in sequence.

[0091] If the parameters of patch GID002 are (112.3531, 24.5682, 20230910143711);

[0092] GID003 is (112.3540, 24.5690, 20230910143900);

[0093] The generated parameter set is:

[0094] {(GID001,112.3512,24.5678,20230910143522),

[0095] (GID002,112.3531,24.5682,20230910143711),

[0096] (GID003,112.3540,24.5690,20230910143900),…};

[0097] This set is used as the patch coding information set for subsequent sorting and grouping.

[0098] The coding sequence construction submodule calls the patch coding information set, sorts all patches according to their acquisition time stamp, reconstructs the patch data sequence in chronological order, identifies the coding differences and spatial adjacency between adjacent patches, groups patches with continuous coding and spatial adjacency, and generates a set of adjacent patch sequences.

[0099] First, call the patch coding information set and sort all patches in ascending order according to the timestamp field. During the execution process, the time stamp "20230910143522" in string format is standardized and converted into a comparable format before being sorted item by item. For example, if the times of patches GID001, GID002, and GID003 are 14:35:22, 14:37:11, and 14:39:00 respectively, the sorting result is GID001→GID002→

[0100] GID003, after sorting is completed, traverse the adjacent spots in turn to determine whether the difference between their spot code values ​​is a unit increment. For example, if the GID suffix numbers are consecutive, the difference is 1. The difference calculation step is to round the two spot ID suffix numbers and then make the difference. For example, GID002-GID001=2-1=1, which meets the coding continuity condition. At the same time, calculate the spatial distance between the center coordinates of the two spots, and use the spherical distance to simplify the calculation formula:

[0101]

[0102] Where Δx is the longitude difference, Δy is the latitude difference, and the unit is degree. For example, GID001(112.3512, 24.5678), GID002(112.3531, 24.5682), then Δx=0.0019, Δy=0.0004,

[0103]

[0104] If the threshold is set to 0.5, it is considered spatially adjacent. If the conditions are met, the patch pair is grouped together. Continuing to perform the above judgment on all sorted patches, multiple matching combination groups can be obtained, such as combination group 1 is {GID001, GID002, GID003}, combination group 2 is {GID005, GID006}, etc. Each group meets the coding continuity and spatial adjacency conditions, and is merged to form a set of adjacent patch sequences.

[0105] The spatial fragment combination submodule extracts the boundary coordinate points of each group of patches based on the set of adjacent patch sequences, calculates the boundary overlap and cross-extension value of each combination of patches, and determines the integrity boundary value of its fragment combination using the formula:

[0106]

[0107] The calculation obtains the combined boundary consistency value, and the corresponding patch combination is selected according to the result that the boundary consistency value is greater than the combined boundary threshold, and its closed boundary path is established to generate the combined boundary structure set;

[0108] Among them, B s represents the combined boundary consistency value, Cz Represents the number of boundary points of the zth patch, L z Represents the length of the boundary line segment of the zth patch, A z Represents the area value of the zth spot, M z Represents the boundary overlap distance between the zth spot and its adjacent spots. Represents the average spatial adjacent distance value of the entire patch group, D u represents the outer boundary extension length of the u-th group of patches, T u Represents the number of patches contained in the u-th group of patches, V u represents the internal boundary jump degree of the u-th group of patches, w represents the total number of patches contained in the currently processed patch combination, and m represents the number of external boundary extension groups involved in the current patch combination;

[0109] After obtaining a set of adjacent patch sequences, we need to calculate the combined boundary consistency value. This process first extracts parameters such as the number of boundary points, boundary segment length, area, and the length of overlap with adjacent patches for each patch group. We then calculate the average adjacency distance, boundary extension length, number of patches, and boundary jump between the combined groups. For the combined group {GID010, GID011, GID012}, the parameters are as follows: GID010 has 35 boundary points, a boundary length of 145 meters, an area of ​​1620 square meters, and a 40-meter overlap with GID011; GID011 has 33 boundary points, a boundary length of 140 meters, an area of ​​1500 square meters, and a 35-meter overlap; and GID012 has 36 boundary points, a boundary length of 150 meters, an area of ​​1580 square meters, and a 30-meter overlap. The average adjacency distance for the combined group is 13.2 meters, boundary extension lengths of 12 meters and 10 meters, respectively, the number of patches is 3, and the boundary jump is 2 and 1, respectively.

[0110] Substitute the above parameters into the consistency formula:

[0111]

[0112] The calculation steps are as follows:

[0113] C1=35,L1=145,M1=40,then: 35·(145+40.25)

[0114] =35·185.25=6483.75, the difference is |6483.75-40|=6443.75;

[0115] C2=33,L2=140,M2=35,then we get: 33·(140+38.73)

[0116] =33·178.73=5898.09, the difference is |5898.09-35|=5863.09;

[0117] C3=36, L3=150, M3=30, get: 36·(150+39.75)

[0118] =36·189.75=6831, the difference is |6831-30|=6801;

[0119] The sum of the numerators is: 6443.75 + 5863.09 + 6801 = 19107.84;

[0120] Denominator: D1=12, T1=3, V1=2; V2=1;

[0121] Total: 13.2 + (6.93 - 2) + (5.77 - 1) = 13.2 + 4.93 + 4.77 = 22.9;

[0122] The denominator is: 3·22.9=68.7;

[0123] Final calculation results:

[0124]

[0125] This result indicates that the boundary consistency of this patch combination is 278.21. If the boundary consistency threshold is set to 250, the judgment criteria are met, indicating that a closed boundary path can be established and the combined boundary structure set of this combination can be output. The formula is beneficial in that it explicitly quantifies the integrity of the spatial structure through the joint calculation of multiple parameters (such as the number of boundary points, the square root of the area, and the deduction of the boundary jump degree), which helps to screen patch combinations with stable boundary structures in big data scenarios.

[0126] See also Figure 5 , the spatial variation stitching module includes:

[0127] The boundary line extraction submodule calls the combined boundary structure set to obtain the original layer content of the vegetation coverage area tile and the wetland distribution area tile, detects the continuous time layer frames of the tile, calls the boundary line data of the tile in the adjacent time frame, compares the boundary direction differences of the tile in the adjacent time frame, and numbers the outline of each frame tile to generate a contour boundary number sequence;

[0128] First, it is necessary to read the continuous remote sensing image layer from May to August 2024, and extract the vegetation tiles and wetland map tiles corresponding to each month from the original layer using coordinate labels. For example, the area with a coordinate range of (100, 200)-(500, 600) is marked as the same area index in each frame of the image to ensure time series consistency. After the extraction is completed, the boundary of each frame of the image block is expressed as a polygon outline. After the outline boundary is vectorized, the boundary nodes on the outline are selected with equal spacing, such as selecting a node every 5 meters to ensure the consistency of the node number. The nodes are represented by two-dimensional coordinates. Then, for each outline node sequence, the angle change between each node is calculated and the boundary direction is marked.

[0129] In the May 2024 frame, the boundary nodes of tile 1 are [(120.3, 210.4), (125.1, 213.7), (130.0, 215.6), (135.5, 214.2)];

[0130] The corresponding tile nodes for June 2024 are [(120.9, 211.1), (125.8, 214.4),

[0131] (130.5, 216.0), (135.9, 214.8)], the contour changes between adjacent time frames are analyzed by node-by-node difference, and the boundary change points are marked at the nodes where the direction angle changes by more than 15 degrees. For the numbering of the contour, the layer time frame index and position index of the tile where the contour is located are double-numbered, such as "202405_03" represents the third tile in May 2024. By performing the above processing on the contours of all time frame tiles, a complete contour boundary numbering sequence is formed, such as [202405_01, 202405_02, …, 202408_01, 202408_02].

[0132] The contour overlap discrimination submodule compares the geometric overlap rates of the boundary lines of adjacent tiles in the continuous layer according to the contour boundary number sequence, calculates the overlap rate value and spatial offset distance value of each pair of adjacent tile contour boundaries, and determines whether it exceeds the set contour offset threshold and spatial position offset threshold using the formula:

[0133]

[0134] The operation obtains the spatial offset value of the block. Based on the situation where the offset threshold is exceeded, each block with boundary drift behavior is obtained, and the starting contour number and the ending contour number are recorded to obtain a list of boundary drift contour intervals.

[0135] Among them, R disp Represents the tile space offset value, L e represents the length of the boundary line of the e-th block, ΔCe Represents the displacement of the contour center between the two frames before and after the e-th block, O e Represents the boundary overlap value of the e-th block, S e represents the average offset of the boundary line of the e-th tile, and g represents the total number of tile contour numbers;

[0136] According to the contour boundary number sequence, the tile space offset value R between consecutive layers disp The calculation can be performed item by item as follows. There are three blocks, namely e1, e2, and e3, and their corresponding boundary line length, contour centroid displacement, boundary coincidence rate, and boundary average offset are:

[0137] Tile e1: L1 = 110 m, ΔC1 = 4.2 m, O1 = 0.68, S1 = 2.0 m;

[0138] Panel e2: L2 = 125 m, ΔC2 = 3.6 m, O2 = 0.61, S2 = 1.9 m;

[0139] Block e3: L3=98m, ΔC3=5.1m, O3=0.52, S3=2.4m;

[0140] Substituting into the original formula:

[0141]

[0142] The sub-items are as follows:

[0143]

[0144] Finally, the three items are summed:

[0145] R disp =|218.7+225.6+203.5|=|647.8|=647.8;

[0146] The offset threshold is set to 400, and the comparison shows that 647.8>400, so it is considered that there is a spatial offset phenomenon in this area, and the marked contour interval list is: [(202405_01)-(202406_01)],

[0147] [(202406_02)-(202407_02)], [(202407_03)-(202408_03)]. By introducing the product of contour length and centroid displacement and combining it with the boundary coincidence rate and average offset to construct a denominator to normalize the offset, the intensity of the effect of tile size on the offset can be quantified. The denominator provides dual control of coincidence accuracy and average offset, making the model highly sensitive to local changes and overall direction changes.

[0148] The tile generation submodule calls the boundary drift contour interval list and selects contour combinations whose spatial overlap exceeds the tile stitching benchmark threshold based on the boundary line continuity between contours and the tile boundary coordinates. It then sorts the contour numbers to form the tile connection order, obtains the tile boundary correspondence, and generates the tile sequence.

[0149] After receiving the boundary drift contour interval list, the mosaic tile generation submodule first performs a node continuity analysis of the spatial boundary of each contour combination. For example, the contour 202405_01 ends at point (300.0, 450.0), and the adjacent contour 202406_01 starts at (300.7, 450.5). The distance between the two points is If the overlap area between the two contours is less than the set continuity threshold of 1.2 meters, it is considered continuous. Then the overlapping area between the two contours is calculated. If the overlapping area between contours 202405_01 and 202406_01 is 43.6 square meters and the joint area of ​​the blocks is 125 square meters, then the overlap rate is Setting the stitching benchmark threshold to 0.35, it can be determined that the current overlap rate is slightly lower than the standard. If the overlapping area of ​​the other two contours 202406_02 and 202407_02 is 58.5 square meters and the joint area is 130 square meters, then the overlap rate is 0.45>0.35, which meets the stitching requirements. All contours that meet the stitching conditions are sorted in ascending order of contour numbers, such as 202406_02<202407_02<202408_02, and the tile connection order is determined to be [202406_02, 202407_02, 202408_02]. According to the corresponding relationship of the nodes on each contour boundary, the coordinates are weighted averaged to generate a unified stitching boundary. For example, the coordinates of the stitching points are:

[0150] The final generated mosaic tile sequence is 202406_02 to 202408_02, numbered "20240608_A".

[0151] See also Figure 6 , the ecological trend mapping module includes:

[0152] The boundary stitching submodule calls the boundary information of each tile in the tile sequence, extracts the spatial contour edge segments, compares the stitching relationship of the edge segments of adjacent tiles, constructs the stitching path, and collects the extension trend of the path in two-dimensional space to generate the tile stitching extension line;

[0153] Based on the boundary information of each tile in the tile sequence, the boundary coordinate set of a single tile must first be extracted. This set usually comes from the vectorized output in the remote sensing data preprocessing stage. For example, the boundary node sequence of tile A01 is [(105.2, 28.4), (106.0, 28.4), (106.0, 29.0), (105.2, 29.0)]. After forming the boundary contour polyline segment, it can be further segmented into an array of line segments with clear directions. The start and end coordinates of each line segment are used as the basic data for determining the splicing relationship. For edge segments across multiple tiles, the Euclidean distance between their endpoints and the starting endpoints of other tiles is sequentially extracted to determine if proximity exists. The direction vectors of the two segments are also extracted, and the angle between them is calculated. Based on this, thresholds for stitching conditions are set: a distance threshold of 15 meters and an angle difference threshold of 20 degrees. These thresholds are derived from a statistical analysis of regional feature scale characteristics, taken as 0.25 times the average minimum boundary dimension of the regional plots, and determined through expert field surveys. For example, if the boundary endpoint of tile A01 is (106.0, 28.4) and the boundary starting point of tile A02 is (106.0, 28.55), the distance between the two points is 11.1 meters, and the angle between them is 13 degrees, meeting the stitching criteria. Therefore, the boundaries of the two tiles are considered continuous. This process processes the boundary information between tiles pairwise, ultimately forming a set of spatially continuous stitching paths. The spatial trend of the splicing path is calculated by calculating the coordinate increments of each line segment. For example, the direction of the line segment from A01 to A02 is due north, the extension distance is 0.15 kilometers, and the path trend vector is (0, +0.15). All the splicing path trend information is summarized and recorded, and finally the tile splicing extension line is obtained.

[0154] The direction recognition submodule calls the tile splicing extension line, extracts the extension path set of multiple time periods, calculates the main axis vector of the path on the spatial coordinate axis, compares the main axis vector direction of different time periods within the same block, and selects the direction distribution line where the direction appears repeatedly for a set number of times to obtain the repeated extension direction value;

[0155] After applying tile splicing extension lines, a unified analysis of path trends over different time periods was performed. The years 2020, 2021, and 2023 were selected as analysis time periods. The extended line segments were uniformly mapped onto a 1km×1km spatial grid, and the direction vector of each line segment within the grid was recorded, expressed as a metric. For example, a line segment extending from (105.2, 28.4) to (105.2, 29.4) has a direction of 0 degrees. Path directions for the same block over different time periods were compared using angle difference calculations. If the direction in 2020 is 5 degrees, 2021 is 7 degrees, and 2023 is 3 degrees, the pairwise differences between the three are 2 degrees, 4 degrees, and 4 degrees, respectively, all within the set directional consistency threshold of 10 degrees. This threshold was set based on the maximum observed fluctuation in main direction during local vegetation change, which does not exceed 10 degrees. This threshold was verified using regional wind direction data and the stable direction of mountain trends. Count the number of occurrences within each spatial grid that meet the directional consistency criteria. If a direction appears more than twice in three time periods, it is considered a repeating direction. In the above example, a trend of approximately 5 degrees appears three times in a row, meeting the criteria. Count all grid blocks that meet the requirements and annotate the direction angle to obtain the repeating extension direction value.

[0156] The trend combination submodule selects line segments with consistent directions in the extension line based on the repeated extension direction values, extracts the spatial block identifiers corresponding to the extension direction, determines the spatial overlap between the distribution lines of the degradation contiguous belts and the location points of the variation concentration areas in the ecological scene, selects line segment combinations with overlaps exceeding the set threshold, and splices them into continuous spatial segments to generate the ecological remote sensing dataset.

[0157] According to the repeated extension direction value, select the line segments with similar directions in the tile splicing extension line. For example, if the repeated direction is 5 degrees, select the tile line segment combination with the direction angle between 2 and 8 degrees, record the corresponding spatial block numbers such as A02, B03, B04, etc., and further extract the boundary vector ranges corresponding to these blocks. Through spatial superposition, perform intersection operation with the degraded contiguous belt vector range to obtain spatial overlapping area data, and then perform ratio operation with the total area of ​​the spatial unit where the tile is located. For example, the intersection area of ​​the B03 area and the degraded belt is 0.48km 2 , with a total area of ​​1.25km 2The calculated overlap ratio was 38.4%, exceeding the set overlap threshold of 30% and therefore retained. This overlap threshold was derived from an analysis of the statistical characteristics of typical degradation areas and the minimum effective coverage ratio for spatially concentrated areas assessed using remote sensing interpretation data from at least five years. The spatial extent of eligible tiles was then compared with the location point data for concentrated variation areas. A buffer analysis was performed, generating a 100-meter-radius buffer zone centered on each location point to determine whether the tile intersected with the buffer zone. For example, if the southern edge of tile B03 intersected with a buffer zone, it was considered a hit point. If two hit points were found, the tile met the criteria for concentrated variation, confirming that B03 possessed multiple overlap characteristics. Finally, all tiles and their corresponding stitching segments that met the criteria for directional consistency, overlap ratio of degradation zones, and location point hits were combined to form a spatially continuous segment, generating the ecological remote sensing dataset.

[0158] The above are merely preferred embodiments of the present invention and do not limit the present invention in any other form. Any technician familiar with the profession may use the technical content disclosed above to change or modify it into an equivalent embodiment with equivalent changes and apply it to other fields. However, any simple modification, equivalent change and modification made to the above embodiment based on the technical essence of the present invention without departing from the content of the technical solution of the present invention shall still fall within the scope of protection of the technical solution of the present invention.

Claims

1. An ecological remote sensing data processing system, characterized by: The system comprises: The remote sensing layer acquisition module obtains the images scanned by the remote sensing platform, combines the image geographic information with the boundary value, identifies the covered objects, matches the known identifiers, constructs the map area, marks the upper, lower, left and right boundary points, and generates a bounding box information set; The boundary position verification module extracts the projection plane position based on the bounding box information set, calls the map reference boundary marking, performs boundary reset according to the offset direction and distance of the diagonal points of the map spot and the reference boundary, relocates the four corners of the map spot, and generates a relocated map spot coding layer; The regional coding integration module calls the relocated patch coding layer, constructs the layer in chronological order, retrieves patches with continuous and adjacent coding, packages them into spatial segments, and generates a combined boundary structure set; The spatial variation stitching module calls the combined boundary structure set, selects vegetation and wetland blocks, compares the overlap of adjacent block outlines, determines boundary drift, records outline numbers, and generates a stitching block sequence; The ecological trend drawing module calls the mosaic tile sequence, retrieves repeated extension directions and expansion forms, matches degradation zones and variation points, combines convergent tiles into trend segments, and obtains an ecological remote sensing dataset.

2. The ecological remote sensing data processing system according to claim 1, characterized in that: The bounding box information set includes the upper left boundary coordinates, lower right boundary coordinates, lower left boundary coordinates, and upper right boundary coordinates of the feature map; the repositioned map coding layer includes a projection boundary plane coordinate set, reference boundary marking space parameters, offset direction identifier, and distance threshold parameters; the combined boundary structure set includes a geocoding unique identifier, a timestamp index, adjacent map spatial topological relationships, and fragment combination closed boundary lines; the spliced ​​tile sequence includes a drift judgment result table, a contour offset matrix, and a junction area splicing path vector set; the ecological remote sensing data set includes the spatial trajectory coordinates of the degraded contiguous belt, the variation concentration area polygon vertex set, the trend fragment vector overlay unit, and the extension direction repeatability identifier.

3. The ecological remote sensing data processing system according to claim 1, characterized in that: The remote sensing layer acquisition module includes: The image geolocation submodule collects ground scanning image data from the remote sensing platform, extracts the geographic identification information carried by each image, compares the corresponding positions of the coordinate points, image time stamps and spatial boundary values ​​in the image, calculates the spatial mapping point distribution of the image data, obtains the location of the image in geographic space, and generates the image spatial positioning coefficient value; The patch boundary recognition submodule calls the area range within the boundary value in the image based on the image spatial positioning coefficient value, calculates the similarity coefficient between the pixel characteristic value and spatial distribution value in the area and the spatial identification of the known ground object, and judges and screens the patches in the area according to the similarity coefficient and the set spatial matching threshold, obtains the spatial partition value of the area covered by the boundary range, and generates a spatial patch matching value; The bounding box generation submodule identifies the contour edge of the covered pattern according to the spatial pattern matching value, determines the upper, lower, left, and right boundary point positions of the edge segment in the image pixel grid, calculates the coordinate values ​​of the boundary point set, and uses the formula: Obtain boundary envelope trend strength values ​​through calculation, construct boundary point sets based on the boundary envelope trend strength values, obtain closed graphic areas of the patch boundaries, and generate boundary box information sets; Among them, B ij Represents the boundary envelope trend strength value, C ik Represents the boundary coordinate difference of the kth point in the i-th area of ​​the edge of the spot contour, D jk Represents the center extension distance of the kth point in the jth area in the edge of the pattern contour, M ijk Represents the weighted matching factor between the boundary point and the matching value, P i Represents the horizontal coordinate difference between the upper boundary point and the center, Q j Represents the vertical coordinate difference between the lower boundary point and the center, and n represents the number of boundary points involved in the calculation in each area.

4. The ecological remote sensing data processing system according to claim 1, characterized in that: The boundary position verification module includes: The boundary point extraction submodule extracts the projection plane boundary position used by the patch based on the boundary points marked in the bounding box information set, identifies the coordinate data and spatial position index corresponding to the boundary points, performs boundary point integrity detection and patch boundary topological relationship judgment, and obtains the boundary coordinate index value; The diagonal offset calculation submodule calls the boundary coordinate index value, selects the diagonal points of the patch corresponding to the reference boundary according to the set of patch boundary points corresponding to the boundary coordinate index value, collects the horizontal and vertical coordinate offset values, analyzes the spatial position change trend between the diagonal points of the patch and the reference mark, and obtains the patch offset angle value set; The boundary position resetting submodule adjusts the four position coordinates of the boundary of the patch according to the offset direction and offset distance of the patch in the patch offset angle value set, performs coordinate difference correction and spatial reprojection operations, and generates a relocated patch coding layer.

5. The ecological remote sensing data processing system according to claim 1, characterized in that: The regional coding integration module includes: The patch coding acquisition submodule calls the relocated patch coding layer to obtain the geographic code and acquisition time mark of each patch, extracts the latitude and longitude coordinate values ​​and timestamp values ​​according to the unique identification value of the patch, and structures the collected parameters using the patch as the index field, constructs a patch parameter set table, and generates a patch coding information set; The coding sequence construction submodule calls the patch coding information set, sorts all patches according to their acquisition time stamps, reconstructs the patch data sequence in chronological order, identifies the coding differences and spatial adjacency between adjacent patches, groups patches with continuous coding and spatial adjacency, and generates a set of adjacent patch sequences. The spatial fragment combination submodule extracts the boundary coordinate points of each group of patches based on the adjacent patch sequence set, calculates the boundary overlap and cross-extension value of the patches, and determines the integrity boundary value of the fragment combination using the formula: The calculation obtains the combined boundary consistency value, and the corresponding patch combination is selected according to the result that the boundary consistency value is greater than the combined boundary threshold, and a closed boundary path is established to generate a combined boundary structure set; Among them, B s represents the combined boundary consistency value, C z Represents the number of boundary points of the zth patch, L z Represents the length of the boundary line segment of the zth patch, A z Represents the area value of the zth spot, M z Represents the boundary overlap distance between the zth spot and its adjacent spots. Represents the average spatial adjacent distance value of the entire patch group, D u represents the outer boundary extension length of the u-th group of patches, T u Represents the number of patches in the uth group, V u represents the internal boundary jump degree of the u-th group of patches, w represents the total number of patches contained in the currently processed patch combination, and m represents the number of external boundary extension groups involved in the current patch combination.

6. The ecological remote sensing data processing system according to claim 1, characterized in that: The space-varying splicing module includes: The boundary line extraction submodule calls the combined boundary structure set, obtains the original layer content of the vegetation coverage area tile and the wetland distribution area tile, detects the continuous time layer frames of the tile, calls the boundary line data of the tile in the adjacent time frames, compares the boundary direction differences of the tile in the adjacent time frames, numbers the outline of each frame tile, and generates a contour boundary number sequence; The contour overlap determination submodule compares the geometric overlap rates of the boundary lines of adjacent tiles in the continuous layer according to the contour boundary number sequence, calculates the overlap rate value and spatial offset distance value of each pair of adjacent tile contour boundaries, and determines whether the set contour offset threshold and spatial position offset threshold are exceeded using the formula: The operation obtains the spatial offset value of the block. Based on the situation where the offset threshold is exceeded, each block with boundary drift behavior is obtained, and the starting contour number and the ending contour number are recorded to obtain a list of boundary drift contour intervals. Among them, R disp Represents the tile space offset value, L e represents the length of the boundary line of the e-th block, ΔC e Represents the displacement of the contour center between the two frames before and after the e-th block, O e Represents the boundary overlap value of the e-th block, S e represents the average offset of the boundary line of the e-th tile, and g represents the total number of tile contour numbers; The mosaic tile generation submodule calls the boundary drift contour interval list, screens contour combinations whose spatial overlap is greater than the tile mosaic reference threshold based on the boundary line continuity between contours and the tile boundary coordinates, sorts the contour numbers to form the tile connection order, obtains the tile boundary correspondence, and generates a mosaic tile sequence.

7. The ecological remote sensing data processing system according to claim 1, characterized in that: The ecological trend drawing module includes: The boundary splicing submodule calls the boundary information of each tile in the splicing tile sequence, extracts the spatial contour edge segments, compares the splicing relationship of the edge segments of adjacent tiles, constructs a splicing path, and collects the extension trend of the path in two-dimensional space to generate a tile splicing extension line; The direction recognition submodule calls the block splicing extension line, extracts the extension path set of multiple time periods, calculates the main axis vector of the path on the spatial coordinate axis, compares the main axis vector directions of different time periods within the same block, and selects the direction distribution lines with repeated directions reaching a set number of times to obtain the repeated extension direction value; The trend combination submodule selects line segments with consistent directions in the extension line according to the repeated extension direction values, extracts the spatial block identifiers corresponding to the extension directions, determines the spatial overlap between the distribution lines of the degradation contiguous belts and the positioning points of the variation concentration areas in the ecological scene, screens the line segment combinations with overlaps exceeding the set threshold, splices them into continuous spatial segments, and generates the ecological remote sensing dataset.

Images