A method and system for measuring the degree of deviation of the succession track of a degraded forest habitat

Abstract

The present application relates to the technical field of data processing and resource management, and specifically discloses a forestry degraded habitat succession track deviation measurement method and system, wherein the present application obtains multi-period three-dimensional laser point cloud data, generates a static physical anchor point set through morphological elevation filtering and density clustering denoising, calculates spatial affine transformation parameters by singular value decomposition with the first period as a benchmark, performs reverse rigid body inverse transformation on the point cloud to generate aligned data, extracts the center of mass of the plant stem to construct an undirected complete graph, obtains the total length of the current habitat structure by running the minimum spanning tree algorithm, and generates a topological structure deviation by comparing with the benchmark value; then, the three-dimensional deviation vector containing water gap, nutrient loss and increased plant spacing is generated by orthogonal decoupling with soil sensing data; finally, the projection coefficient is solved by linear equation system, the accurate deployment list of irrigation water, compound fertilizer and seedlings is generated, and the automatic terminal is executed to realize closed-loop intelligent management.

Description

Technical Field

[0001] This invention belongs to the field of data processing and resource management technology, and relates to a method and system for calculating the offset of succession trajectory of forest degraded habitats. Background Technology

[0002] In modern forestry management, forests are not only an important component of the ecosystem but also considered dynamic assets with economic and social value. To effectively regulate and enhance these assets, forestry management is gradually shifting from a traditional experience-driven model to a data-driven, intelligent management model. This transformation relies on continuous monitoring and accurate assessment of forest habitat health, as well as scientific resource allocation and intervention decisions based on the assessment results.

[0003] In existing technologies, monitoring forest habitats typically employs a variety of separate technical methods. On one hand, satellite remote sensing or drone aerial photography is used to acquire macroscopic two-dimensional image information such as vegetation cover and normalized difference vegetation index (NDVI) over a wide area, which is used to preliminarily assess the overall growth status of the forest land. On the other hand, environmental parameters such as soil moisture and nutrient content are obtained at specific locations by deploying sensors within the forest land or by conducting manual sampling. Based on these data from different sources and with varying dimensions, combined with forestry knowledge and personal experience, managers comprehensively assess whether the forest land is at risk of degradation and formulate corresponding maintenance plans accordingly.

[0004] However, the aforementioned existing technical solutions have inherent shortcomings in practical applications. First, the lack of an effective quantitative correlation model between macroscopic remote sensing data and microscopic soil data results in the diagnosis of habitat degradation remaining at a qualitative or semi-quantitative level, making it difficult to accurately identify the root causes of degradation. Second, there is a significant gap between monitoring results and management measures. Diagnostic reports typically cannot directly generate a clear list of resource requirements. The transformation from "identifying the problem" to "how much water and fertilizer is needed" relies on manual estimation, lacking objectivity and standardization, leading to insufficient accuracy in resource allocation and frequent instances of over- or under-intervention. Summary of the Invention

[0005] In view of this, in order to solve the problems mentioned in the background technology, a method and system for calculating the offset of the succession trajectory of forest degraded habitats is proposed.

[0006] The objective of this invention can be achieved through the following technical solutions: The first aspect of this invention provides a method for calculating the trajectory offset of forest degradation habitat succession, including: S1, acquiring three-dimensional laser point cloud physical data of target forest land under different sampling periods, performing morphological elevation filtering, extracting a set of three-dimensional physical coordinate points of vegetation base that are close to the ground and unaffected by wind disturbance, and generating a set of static physical anchor points.

[0007] S2. Using the set of stationary physical anchor points in the first sampling period as the absolute reference, the singular value decomposition algorithm is used to calculate the rotation matrix and translation matrix of the set of stationary physical anchor points in any subsequent sampling period relative to the absolute reference, and spatial affine transformation parameters are generated.

[0008] S3. Using spatial affine transformation parameters, perform inverse rigid body transformation on the physical data of the three-dimensional laser point cloud for any subsequent sampling period to generate deformation-free three-dimensional point cloud data with strictly aligned physical coordinates.

[0009] S4. Extract the centroid 3D physical coordinates of all vegetation trunks from the non-deformable 3D point cloud data, use the centroid 3D physical coordinates as discrete nodes, and use the Euclidean physical space distance between discrete nodes as the edge length weight to construct an undirected complete graph.

[0010] S5. Run the minimum spanning tree algorithm on the undirected complete graph to calculate the minimum spanning tree connecting all discrete nodes of the target forest, and sum the topological edge lengths of the minimum spanning tree to generate the total length of the current habitat structure.

[0011] S6. Obtain the total length of the baseline habitat structure of the target forest land under standard healthy succession, calculate the mathematical difference ratio between the current total habitat structure length and the baseline total habitat structure length, and generate the topological offset.

[0012] S7. Obtain basic soil sensing data of the target forest land, and perform orthogonal decoupling processing on the topological structure offset and basic soil sensing data to generate a three-dimensional physical deviation vector containing soil moisture deficit, nutrient loss equivalent and vegetation spacing increase.

[0013] S8. Obtain the physical resource cost basis matrix configured in the management system, calculate the projection coefficient of the three-dimensional physical deviation vector on the physical resource cost basis matrix by solving the linear equation system, generate a physical material allocation list based on the projection coefficient, and send the physical material allocation list to the automated repair execution terminal for entity resource scheduling.

[0014] The second aspect of the present invention provides a system for calculating the trajectory offset of forest degradation habitat succession, comprising: a static physical anchor point set generation module, which acquires three-dimensional laser point cloud physical data of target forest land under different sampling periods, performs morphological elevation filtering, extracts a set of three-dimensional physical coordinate points of vegetation base that are close to the ground and unaffected by wind disturbance, and generates a static physical anchor point set.

[0015] The spatial affine transformation parameter generation module uses the set of stationary physical anchor points in the first sampling period as the absolute reference and employs the singular value decomposition algorithm to calculate the rotation and translation matrices of the set of stationary physical anchor points in any subsequent sampling period relative to the absolute reference, thereby generating spatial affine transformation parameters.

[0016] The non-deformable 3D point cloud data generation module uses spatial affine transformation parameters to perform an inverse rigid body transformation on the physical data of the 3D laser point cloud in any subsequent sampling period, generating non-deformable 3D point cloud data with strictly aligned physical coordinates.

[0017] The undirected complete graph construction module extracts the centroid 3D physical coordinates of all vegetation trunks from the unvariable 3D point cloud data, uses the centroid 3D physical coordinates as discrete nodes, and uses the Euclidean physical space distance between discrete nodes as the edge length weight to construct an undirected complete graph.

[0018] The current habitat structure total length generation module runs the minimum spanning tree algorithm on the undirected complete graph, calculates the minimum spanning tree connecting all discrete nodes of the target forest, and sums up the topological edge lengths of the minimum spanning tree to generate the current habitat structure total length.

[0019] The topology offset generation module obtains the total length of the baseline habitat structure when the target forest is in a standard healthy succession state, calculates the mathematical difference ratio between the current total habitat structure length and the baseline total habitat structure length, and generates the topology offset.

[0020] The three-dimensional physical deviation vector generation module acquires the basic soil sensing data of the target forest land, performs orthogonal decoupling processing on the topological structure offset and the basic soil sensing data, and generates a three-dimensional physical deviation vector containing soil moisture deficit, nutrient loss equivalent, and vegetation spacing increase.

[0021] The material list mapping and scheduling module obtains the physical resource cost basis matrix configured in the management system, calculates the projection coefficient of the three-dimensional physical deviation vector on the physical resource cost basis matrix by solving a system of linear equations, generates a physical material allocation list based on the projection coefficient, and sends the physical material allocation list to the automated repair execution terminal for entity resource scheduling.

[0022] Compared with the prior art, the embodiments of the present invention have at least the following advantages or beneficial effects: (1) The present invention optimizes the forestry management model by constructing a complete technical link from three-dimensional physical data acquisition to automated resource scheduling. It deeply integrates high-precision three-dimensional laser point cloud data with ground sensor data, and transforms the complex ecosystem degradation problem into quantifiable mathematical indicators by establishing a topological model of habitat structure. This method is different from the traditional management method that relies on macroscopic image interpretation or discrete sample point surveys, and provides an objective, detailed and traceable evaluation basis for the health status of forestry assets, enabling management decisions to shift from post-event remediation to pre-event early warning and precise intervention.

[0023] (2) This invention proposes an orthogonally decoupled diagnostic framework that can decompose comprehensive habitat degradation indicators into independent physical resource gaps. By combining topological offset with soil sensing data for analysis, the system can clearly distinguish the degradation contributions caused by different factors such as changes in vegetation spatial patterns, soil moisture imbalance, or nutrient loss. This diagnostic approach avoids the "one-size-fits-all" restoration measures in traditional management, making resource input more targeted and improving the efficiency and effectiveness of restoration activities such as irrigation, fertilization, and replanting, which meets the core requirements of sustainable resource management.

[0024] (3) This invention ultimately integrates ecological diagnostic results with physical resource scheduling, forming a closed-loop automated management system. By introducing a physical resource cost basis matrix, the calculated three-dimensional physical deviation vector is directly transformed into an allocation list containing specific material quantities, and instructions that can be executed by automated equipment are generated. This not only reduces the resource input for manual inspection, decision-making, and operation, but also shortens the response cycle from problem discovery to solution, realizing dynamic, efficient, and intelligent management of forestry ecological assets. Attached Figure Description

[0025] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the 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.

[0026] Figure 1 This is a schematic diagram of the method steps of the present invention.

[0027] Figure 2 This is a schematic diagram of the system structure connection of the present invention. Detailed Implementation

[0028] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0029] Please see Figure 1 The first aspect of the present invention provides a method for calculating the trajectory offset of forest degradation habitat succession, including: S1, acquiring three-dimensional laser point cloud physical data of target forest land under different sampling periods, performing morphological elevation filtering processing, extracting a set of three-dimensional physical coordinate points of vegetation base that are close to the ground and unaffected by wind disturbance, and generating a set of static physical anchor points.

[0030] In a specific embodiment of the present invention, generating a set of static physical anchor points includes: acquiring three-dimensional laser point cloud physical data of the target forest at different sampling periods.

[0031] Morphological elevation filtering is performed on the physical data of 3D laser point clouds to separate ground point clouds from non-ground point clouds.

[0032] Point cloud data with elevation values ​​within the near-ground threshold range are extracted from non-ground point clouds and used as a set of three-dimensional physical coordinate points for the vegetation base.

[0033] Density clustering is used to denoise the three-dimensional physical coordinate point set at the base of vegetation, eliminating isolated noise points and generating a set of static physical anchor points.

[0034] Specifically, this step aims to accurately extract reference points representing the stable structure of the forest from the raw laser scanning data. First, it is necessary to acquire 3D laser point cloud physical data generated by scanning the same target forest at different sampling periods. The 3D laser point cloud physical data is a large dataset where each data point records the 3D spatial coordinates of the laser pulse as it travels from the device to the surface of the object and back. ,in These represent the physical location of the point in the Cartesian coordinate system. Different sampling periods mean that the forest land is scanned multiple times over time to capture its dynamic changes.

[0035] After acquiring the 3D laser point cloud physical data, morphological elevation filtering is performed on this data. The purpose of this processing is to separate the point cloud into ground point clouds and non-ground point clouds. The core idea of ​​morphological elevation filtering is to simulate a virtual surface with a certain rigidity, "covering" the entire point cloud data from below. Points that can be touched by this virtual surface are identified as ground points, while the rest are non-ground points, such as vegetation and buildings. Through this step, we obtain a set containing only ground points, i.e., the ground point cloud, and a set containing all vegetation, shrubs, and other objects above the ground, i.e., the non-ground point cloud.

[0036] Next, from the non-ground point cloud generated in the previous step, point cloud data with elevation values ​​within the near-ground threshold range are extracted. These points are considered the stable base of vegetation because they are close to the ground and least affected by swaying disturbances caused by environmental factors such as wind. Here, "elevation value" does not refer to the absolute elevation of the point. The coordinates are not the coordinates, but rather the relative height of the point relative to the local ground. For any point in a non-ground point cloud... Its relative height The calculation method is as follows:

[0037] In the formula, Is this non-ground point Coordinate values. Is this point at? The local ground elevation value corresponding to the planar projection location can be obtained by finding the nearest neighbor point in the ground point cloud or by performing local planar interpolation. The filtering criteria are used to determine... Whether it falls within the preset near-ground threshold range, i.e. . The setting was 0.1 meters to exclude interference from weeds and fallen leaves close to the ground surface, and this setting was based on general observations of forest understory cover. The depth was set to 1.0 meter to ensure that only the base of the vegetation trunk was extracted, avoiding the inclusion of easily swaying middle and upper branches. This setting was based on statistical analysis of the base morphology of common tree species. All point cloud data that met this condition constituted the three-dimensional physical coordinate point set of the vegetation base.

[0038] Finally, to further refine the data, density clustering denoising was performed on the three-dimensional physical coordinate point set at the base of the vegetation. This process effectively removes isolated noise points caused by measurement errors or airborne fluff. Density clustering denoising, such as density-based spatial clustering algorithms, works by identifying high-density regions in the data space. It requires two key parameters: neighborhood radius... and minimum points If a point The neighborhood contains no fewer than If a point is reachable from a core point, then that point is defined as a core point. The set of all points reachable from a core point forms a cluster. A point that does not belong to any cluster, i.e., is not a point at any core point, is a point that is not a core point. Sparse points outside the neighborhood are identified as isolated noise points and discarded. In this embodiment, the neighborhood radius... Set to 0.15 meters, minimum number of points. The parameter setting of 10 was determined based on statistical analysis of 200 sets of measured data from industrial-grade LiDAR in a forest environment, ensuring the preservation of the actual vegetation base point cloud and effective noise filtering. After this denoising process, the remaining point cloud set is the final static physical anchor point set. The points in this set have high spatial stability and temporal consistency, serving as the foundation for accurate registration of multi-phase point cloud data in subsequent steps.

[0039] For example, assume that three-dimensional laser point cloud physical data of the target forest was acquired in the first sampling period, which includes the following representative data points: point A (2.1, 3.5, 0.15), point B (2.2, 3.6, 0.18), point C (5.8, 6.2, 3.45), point D (5.9, 6.1, 0.85), point E (5.85, 6.15, 0.95), point F (5.75, 6.25, 0.75), and point G (10.5, 11.2, 0.60).

[0040] First, morphological elevation filtering was performed on all acquired 3D laser point cloud physical data. After processing, points A and B were identified as ground point clouds due to their low elevation values, while points C, D, E, F, and G were identified as non-ground point clouds.

[0041] Next, point cloud data with elevation values ​​within the near-ground threshold range are extracted from the non-ground point cloud. A lower limit for the near-ground threshold range is set. It is 0.1 meters, with an upper limit. The value is 1.0 meter. Assume the local ground elevation at points C, D, E, F, and G is 1.0 meter. All are 0.20 meters. For point C (5.8, 6.2, 3.45), its relative height is... Rice, not enough The condition is met, therefore it is excluded. For point D(5.9, 6.1, 0.85), its relative height... The value in meters satisfies the condition and is retained. For point E(5.85, 6.15, 0.95), its relative height is... The value in meters satisfies the condition and is retained. For point F(5.75, 6.25, 0.75), its relative height is... The value in meters satisfies the condition and is retained. For point G(10.5, 11.2, 0.60), its relative height is... The points that meet the conditions are retained. After this step, the generated three-dimensional physical coordinate set of the vegetation base is {point D, point E, point F, point G}.

[0042] Finally, density clustering was performed on the three-dimensional physical coordinate point set at the base of the vegetation to remove noise. A neighborhood radius was set. The minimum number of points is 0.15 meters. The minimum number of points is 3. Points D, E, and F are very close to each other in three-dimensional space, all within a 0.15-meter neighborhood radius. Taking point D as an example, its neighborhood includes points E and F, plus itself, for a total of 3 points, satisfying the minimum number of points. To meet the requirement of 3, points D, E, and F together form a high-density cluster. However, the physical spatial distance between point G (10.5, 11.2, 0.60) and points D, E, and F is much greater than 0.15 meters, and there are no other points in its neighborhood. Therefore, it does not meet the minimum number of points requirement and is identified as an isolated noise point and is removed.

[0043] Finally, the generated set of static physical anchor points is {point D(5.9,6.1,0.85), point E(5.85,6.15,0.95), point F(5.75,6.25,0.75)}.

[0044] S2. Using the set of stationary physical anchor points in the first sampling period as the absolute reference, the singular value decomposition algorithm is used to calculate the rotation matrix and translation matrix of the set of stationary physical anchor points in any subsequent sampling period relative to the absolute reference, and spatial affine transformation parameters are generated.

[0045] In a specific embodiment of the present invention, generating spatial affine transformation parameters includes: extracting the set of stationary physical anchor points in the first sampling period as an absolute reference, and extracting the set of stationary physical anchor points in any subsequent sampling period as a set of points to be registered.

[0046] Calculate the centroid coordinates of the absolute datum and the centroid coordinates of the point set to be registered, and then perform centroid removal processing on the absolute datum and the point set to be registered to generate the centroid-removed point set of the datum and the centroid-removed point set of the point set to be registered.

[0047] Construct the covariance matrix between the baseline decentroid set and the decentroid set to be registered, and process the covariance matrix using the singular value decomposition algorithm to generate the rotation matrix.

[0048] The translation matrix is ​​calculated based on the rotation matrix, the centroid coordinates of the absolute reference, and the centroid coordinates of the set of points to be registered. The rotation matrix and the translation matrix are then combined to generate spatial affine transformation parameters.

[0049] Specifically, the core task of this step is to accurately calculate the spatial pose differences between point cloud data acquired at different time points, thereby providing a foundation for subsequent data alignment. In step S1, we have already generated a set of stationary physical anchor points for each sampling period. The points in these sets serve as ideal references due to the stability of their physical positions.

[0050] First, the set of stationary physical anchor points in the first sampling period is established as the absolute reference, denoted as set. , which includes Three-dimensional coordinate points Simultaneously, the set of stationary physical anchor points in any subsequent sampling period is taken as the set of points to be registered, denoted as set. This also includes Individual and The three-dimensional coordinates corresponding to the middle anchor point These two sets are the direct inputs for subsequent calculations.

[0051] To eliminate the interference of translation on attitude calculation, both the absolute reference and the point set to be registered need to be decentroided. This process first calculates the centroid coordinates of the two point sets. The centroid coordinates are the arithmetic mean of the three-dimensional coordinates of all points in the point set, representing the geometric center of the point set. The centroid coordinates of the absolute reference... Centroid coordinates of the set of points to be registered The calculation formula is as follows:

[0052]

[0053] In the formula, It is the first in the absolute benchmark The three-dimensional coordinate vector of a point It is the first in the set of points to be registered. The three-dimensional coordinate vector of a point This represents the total number of points in the set of stationary physical anchor points. After calculating the centroid coordinates, the reference centroid set is generated by subtracting the corresponding centroid coordinates from the coordinates of each point set. and the set of points to be matched - the set of centroids to be determined Processed points and The calculation method is as follows:

[0054]

[0055] This operation is equivalent to moving the origin of the coordinate system of both point sets to their respective centroids, allowing subsequent calculations to focus on solving the rotation relationship.

[0056] Next, we construct the covariance matrix of the baseline decentrifuged point set and the decentrifuged point set to be registered. The covariance matrix describes the linear dependence between two decentroided point sets, and its structure implies spatial matching rotation information between the two point sets. It is a 3x3 matrix, and its calculation formula is:

[0057] In the formula, It is a column vector of point coordinates from the reference centroid set. It is the row vector of point coordinates in the centroid set to be registered. For the covariance matrix... Singular value decomposition (SVD) is used to process the matrix, a mathematical method that decomposes a matrix into three special matrices, revealing the matrix's principal components. The decomposition form is as follows:

[0058] In the formula, and They are 3x3 orthogonal matrices, and their column vectors form an orthonormal basis. It is a 3x3 diagonal matrix, where the elements on the diagonal are singular values, representing the scaling of the data along the principal directions. This is obtained through decomposition. and A matrix can be used to calculate the rotation matrix that rotates the set of points to be registered to be aligned with the absolute reference space. :

[0059] In very rare cases, data noise may cause the calculated... This is a transformation matrix that includes mirror reflection, and its determinant is -1. To ensure a purely physical rotation transformation, corrections are needed. If... Then Multiply the last column of the matrix by -1 and then recalculate. To ensure .

[0060] Finally, based on the obtained rotation matrix Centroid coordinates of absolute reference Centroid coordinates of the set of points to be registered Calculate the translation matrix The translation matrix is ​​actually a three-dimensional translation vector that describes how much the centroid of the set of points to be registered needs to be moved to coincide with the centroid of the absolute reference after rotation and alignment. Its relationship can be expressed as follows: From this, the translation matrix can be derived. The calculation formula is as follows:

[0061] Finally, the calculated rotation matrix Translation matrix Combine these parameters to generate spatial affine transformation parameters. This set of parameters fully describes the rigid transformation relationship of the point cloud in any subsequent sampling period relative to the point cloud in the first sampling period.

[0062] For example, suppose that in step S1, a set of static physical anchor points containing three corresponding points is generated from the first sampling period and the second sampling period respectively.

[0063] The set of stationary physical anchor points in the first sampling period serves as the absolute reference. Its coordinates are: , , The set of stationary physical anchor points in the second sampling period is used as the set of points to be registered. Its coordinates are: , , .

[0064] First, calculate the centroid coordinates of the absolute reference. Centroid coordinates of the set of points to be registered : , .

[0065] Next, the two point sets are decentroided to generate a reference decentroided point set. And the set of centroids to be registered : , .

[0066] , , , .

[0067] Then, construct the covariance matrix. The calculation result is: For the covariance matrix Perform singular value decomposition algorithm to obtain the matrix. and The decomposition results are given directly here: , .

[0068] Calculate the rotation matrix : .

[0069] Finally, based on the rotation matrix Calculate the translation matrix using the coordinates of the two centroids. : .

[0070] The final generated spatial affine transformation parameters are derived from the rotation matrix. Translation matrix Together, they precisely describe the spatial transformation relationship of the set of stationary physical anchor points in the second sampling period relative to the first sampling period.

[0071] S3. Using spatial affine transformation parameters, perform inverse rigid body transformation on the physical data of the three-dimensional laser point cloud for any subsequent sampling period to generate deformation-free three-dimensional point cloud data with strictly aligned physical coordinates.

[0072] In a specific embodiment of the present invention, generating non-deformable 3D point cloud data with strictly aligned physical coordinates includes: acquiring each original 3D data point in the 3D laser point cloud physical data of any subsequent sampling period to be tested.

[0073] Each original 3D data point is multiplied by the rotation matrix in the spatial affine transformation parameters to generate rotated and aligned data points.

[0074] The rotation-aligned data points are generated by performing vector addition with the translation matrix in the spatial affine transformation parameters.

[0075] Perform matrix multiplication and vector addition operations on all original 3D data points in the 3D laser point cloud physical data of any subsequent sampling period to generate deformation-free 3D point cloud data with strictly aligned physical coordinates.

[0076] Specifically, after calculating the spatial affine transformation parameters in step S2, the goal of this step is to transform the physical data of the entire three-dimensional laser point cloud for any subsequent sampling period to the coordinate system of the first sampling period, achieving precise spatial overlap of data from different periods. This process is called inverse rigid body transformation, which ensures that the comparison and analysis between point cloud data are performed on a unified and stable spatial reference, eliminating systematic errors introduced by minor changes in the equipment's installation position or attitude.

[0077] This transformation process iterates through every original 3D data point in the 3D laser point cloud physical data of any subsequent sampling period. For any original 3D data point, its coordinates can be represented as a 3D column vector. First, the original 3D data points are compared with the rotation matrix in the spatial affine transformation parameters generated in step S2. Perform matrix multiplication. Rotate the matrix. It is a 3x3 orthogonal matrix that defines the rotation relationship from the coordinate system of any subsequent sampling period to the coordinate system of the first sampling period. This operation rotates the coordinate vector of the original point so that its direction is aligned with the coordinate system of the absolute reference, thus generating an intermediate result, namely, rotated and aligned data points. The mathematical expression for this calculation process is:

[0078] In the formula, It is a three-dimensional coordinate vector after rotation transformation. It is a 3x3 rotation matrix. It is the coordinate vector of the original 3D data points. The essence of this operation is to transform the vector... The basis vectors in the original coordinate system are decomposed and then recombined using the basis vectors in the new coordinate system to obtain the representation in the new coordinate system.

[0079] Next, the rotation and alignment data points generated in the previous step are... Translation matrix in spatial affine transformation parameters Perform vector addition. Translation matrix. This is a three-dimensional column vector representing the physical displacement between the origins of two coordinate systems. By adding the rotated point vector to this translation vector, the coordinate systems can be aligned through translation. The result of this addition operation is the aligned data points. It is the original three-dimensional data point. The final coordinates in the coordinate system of the first sampling period. The calculation formula is:

[0080] Combining the two calculation steps above, for any original 3D data point in any subsequent sampling period to be measured... The complete transformation from the target coordinate system to the target coordinate system can be uniformly represented as:

[0081] In the formula, These are the coordinates of the three-dimensional data points obtained after transformation, which are strictly aligned with the physical coordinates of the data from the first sampling period. and These are the previously calculated rotation and translation matrices, respectively. These are the coordinates of the original 3D data points to be transformed. By repeatedly performing this transformation operation, consisting of matrix multiplication and vector addition, on all the original 3D data points in the 3D laser point cloud physical data of any subsequent sampling period, a completely new point cloud dataset is generated. All points in this dataset are located in the coordinate system of the first sampling period, thus forming a physically coordinate-aligned, deformation-free 3D point cloud data. The term "deformation-free" here emphasizes that the transformation is rigid, meaning that the relative distances and angles between points remain unchanged before and after the transformation; only the overall spatial position and orientation of the point cloud are altered.

[0082] For example, following the example from step S2, we have obtained the spatial affine transformation parameters connecting the second sampling period coordinate system and the first sampling period coordinate system, including the rotation matrix. Translation matrix : , .

[0083] Now, from the 3D laser point cloud physical data of the second sampling period, obtain an original 3D data point that was not involved in the S2 calculation, such as a point on a tree canopy, with coordinates as follows: .

[0084] First, this original 3D data point With rotation matrix Perform matrix multiplication to generate rotated and aligned data points. : Next, the generated rotation-aligned data points are... With translation matrix Perform vector addition to generate the final translation-aligned data points. : .

[0085] Therefore, the original three-dimensional data points After the inverse rigid body transformation, its new coordinates in the coordinate system of the first sampling period are: By performing the same matrix multiplication and vector addition operations on all points in the 3D laser point cloud physical data of the second sampling period, the final generated data set is the physically aligned, deformation-free 3D point cloud data.

[0086] S4. Extract the centroid 3D physical coordinates of all vegetation trunks from the non-deformable 3D point cloud data, use the centroid 3D physical coordinates as discrete nodes, and use the Euclidean physical space distance between discrete nodes as the edge length weight to construct an undirected complete graph.

[0087] In a specific embodiment of the present invention, constructing an undirected complete graph includes: performing cylindrical model fitting processing on the non-variable 3D point cloud data to identify the vegetation trunk point cloud set in the non-variable 3D point cloud data.

[0088] Calculate the geometric center coordinates of the vegetation trunk point cloud set in three-dimensional space, and use the geometric center coordinates as the three-dimensional physical coordinates of the centroid of the vegetation trunk.

[0089] Define the three-dimensional physical coordinates of the centroid as discrete nodes in the graph theory model, and calculate the Euclidean physical space distance between any two discrete nodes.

[0090] Using Euclidean physical distance as the edge length weight connecting two corresponding discrete nodes, we connect all discrete nodes to construct an undirected complete graph.

[0091] Specifically, this step aims to abstract the aligned, continuous 3D point cloud data from the previous step, transforming it into a discrete mathematical structure capable of describing the spatial distribution of vegetation in the forest. This process begins with identifying and extracting the core structure representing a single plant, namely the vegetation trunk, from the physically aligned, non-variable 3D point cloud data.

[0092] To identify vegetation trunks, a cylindrical model fitting process is performed on the deformation-free 3D point cloud data. This process is a geometric shape recognition technique that works by searching the point cloud for point sets that conform to the geometric features of a cylinder. The process iteratively attempts to fit hypothetical cylinder model parameters using different point sets as seeds, including the cylinder's axial direction, the position of a point on the axis, and the cylinder's radius. Then, it evaluates how many points in the entire point cloud data can match this hypothetical cylinder model within a set error tolerance. Through iterative optimization, the subset of point clouds that best represents the true vegetation trunk is finally found. Points successfully identified as belonging to a certain cylinder model are categorized into a vegetation trunk point cloud set. Multiple trees in a forest will correspondingly result in the identification of multiple independent vegetation trunk point cloud sets.

[0093] After obtaining the point cloud set representing each vegetation trunk, a unique representative point needs to be calculated for each set, namely the three-dimensional physical coordinates of the centroid of the vegetation trunk. This coordinate is obtained by calculating the coordinates of the geometric centers of all points constituting the set in three-dimensional space. For the... Collection of vegetation trunk point clouds It contains Three-dimensional points The three-dimensional physical coordinates of the centroid of its main vegetation trunk The calculation formula is:

[0094] In the formula, It is the first The three-dimensional physical coordinate vector of the centroid of a vegetation trunk. It is the first The total number of points in a vegetation trunk point cloud set. It is the th in this set The three-dimensional coordinate vector of the centroid. This three-dimensional physical coordinate of the centroid can be regarded as a highly condensed and representative physical location of the main trunk of the vegetation in three-dimensional space.

[0095] Next, the three-dimensional physical coordinates of the centroids of all the main trunks of vegetation in the forest are used as discrete nodes in the graph theory model. Graph theory is a branch of mathematics that uses nodes and edges to represent objects and their relationships. Here, each discrete node represents a real vegetation tree. To describe the spatial relationships between these vegetation trees, the Euclidean physical distance between any two discrete nodes needs to be calculated. For any two discrete nodes, their three-dimensional physical coordinates of the centroids are respectively... and The Euclidean physical distance between them The calculation formula is:

[0096] In the formula, It is a connection node and nodes The straight-line distance, whose dimensions are consistent with those of the coordinate system, is, for example, a meter. This distance directly reflects the physical distance between two vegetation plants.

[0097] Finally, the calculated Euclidean physical distance is used as the edge weight connecting the corresponding two discrete nodes, and all discrete nodes are connected to construct an undirected complete graph. An undirected complete graph is a special graph structure in which there is an edge between any two distinct nodes, and the edge has no direction. In this graph, nodes represent vegetation, and the edge weight connecting any two nodes is the actual physical distance between those two vegetation trees. This undirected complete graph comprehensively and quantitatively describes the topological spatial distribution pattern of all vegetation trunks in the target forest.

[0098] For example, assuming that in the physically coordinate-aligned, non-deformable 3D point cloud data generated in step S3, two vegetation trunk point cloud sets are identified through cylindrical model fitting processing, and named Vegetation Trunk Point Cloud Set 1 and Vegetation Trunk Point Cloud Set 2 respectively.

[0099] Vegetation trunk point cloud set 1 contains 3 points with the following coordinates: , , Vegetation trunk point cloud set two contains 3 points with coordinates as follows: , , .

[0100] First, calculate the 3D physical coordinates of the centroid of the vegetation trunks in both vegetation trunk point cloud sets. For vegetation trunk point cloud set one, its 3D physical coordinates of the centroid are... for: For the vegetation trunk point cloud set two, its centroid's three-dimensional physical coordinates... for: .

[0101] Next, the three-dimensional physical coordinates of these two centroids will be... and Let node 1 and node 2 be two discrete nodes in a graph theory model. Then, calculate the Euclidean physical distance between these two discrete nodes. : rice.

[0102] Finally, this Euclidean physical distance The weight of the edge connecting node 1 and node 2 is used. Since there are only two nodes, the constructed undirected complete graph contains node 1 and node 2, and an edge connecting them with a weight of 6.948. This graph structure represents the spatial relationship between the two vegetation plants.

[0103] S5. Run the minimum spanning tree algorithm on the undirected complete graph to calculate the minimum spanning tree connecting all discrete nodes of the target forest, and sum the topological edge lengths of the minimum spanning tree to generate the total length of the current habitat structure.

[0104] In a specific embodiment of the present invention, generating the total length of the current habitat structure includes: arranging all edges in the undirected complete graph in ascending order according to the edge length weights from smallest to largest to generate an ordered set of edges.

[0105] Select the edge with the smallest weight from the ordered edge set in turn, and determine whether the selected edge forms a cycle in the selected edge set.

[0106] If no loop is formed, the selected edge is added to the minimum spanning tree edge set until the number of edges in the minimum spanning tree edge set is equal to the total number of discrete nodes minus one, thus generating a minimum spanning tree connecting all discrete nodes of the target forest.

[0107] The weights of the edge lengths of all edges in the minimum spanning tree edge set are summed to generate the total length of the current habitat structure.

[0108] Specifically, after constructing the undirected complete graph describing the spatial distribution of vegetation in step S4, this step aims to extract the core skeleton structure from this complex fully connected network and quantify it into a single index. This process is achieved by running the minimum spanning tree algorithm, ultimately obtaining a value that represents the connectivity and compactness of the current forest habitat, i.e., the total length of the current habitat structure.

[0109] First, all edges in the undirected complete graph are sorted in ascending order according to their edge length weights assigned in step S4, which are the Euclidean physical distances between vegetation nodes. This operation produces an ordered set of edges that connect all discrete node pairs, and these edges are arranged in ascending order of their physical length.

[0110] Next, a greedy strategy is adopted to sequentially select the edge with the smallest weight from the ordered edge set. A crucial judgment is needed after each edge selection: whether adding the newly selected edge to the already selected edge set will form a cycle. A cycle is a closed path consisting of nodes and edges. The method for determining whether a cycle is formed is as follows: initially, each discrete node is considered an independent connected component; whenever an edge is added, if the two nodes connected by this edge belong to two different connected components, a cycle will not be formed, and these two connected components can be merged into one; conversely, if the two nodes already belong to the same connected component, adding this edge will inevitably form a cycle, and therefore this edge must be discarded.

[0111] If the result of the judgment is that no cycle is formed, the selected edge is formally added to a new set, which is called the minimum spanning tree edge set. This process is repeated continuously, selecting the next shortest edge from the ordered edge set and checking for cycles, until the number of edges in the minimum spanning tree edge set is equal to the total number of discrete nodes minus one. For a graph with discrete nodes, a structure connecting all nodes and containing no loops (i.e., a tree structure) must contain exactly the following: The minimum spanning tree edge set and its associated nodes together constitute the minimum spanning tree connecting all discrete nodes in the target forest. This tree is the backbone network connecting all vegetation using the shortest total path.

[0112] Finally, the edge length weights of all edges in the minimum spanning tree edge set are summed. This sum represents the shortest total path length required to connect all vegetation trunks in the forest. This value is defined as the total length of the current habitat structure, using... It is expressed as follows. The calculation formula is:

[0113] In the formula, It is the total length of the current habitat structure; It is the set of edges of the minimum spanning tree; It is any edge in that set; It is the edge The side length weight corresponds to the Euclidean physical distance between the centroids of the two vegetation trunks. The dimensions of this formula are consistent with those of the side length weight, both being units of length, such as meters.

[0114] For example, continuing from the example of step S4, we add a discrete node to better illustrate this step. Assume that three vegetation trunks are identified through step S4, and their centroid 3D physical coordinates are used as discrete nodes, namely: Node 1: Node 2: Node 3: .

[0115] First, constructing an undirected complete graph requires calculating the Euclidean physical distance between all pairs of nodes as edge weights. The edge weight for edge (1,2) is given below. Meters (see example S4 for the calculation process).

[0116] Edge weight of edge (1,3) Meters. The weight of the edge (2,3). rice.

[0117] Next, all edges in the undirected complete graph are sorted in ascending order of their weights, generating an ordered edge set: edge (1,3), weight 4.423; edge (2,3), weight 4.538; edge (1,2), weight 6.948. Then, edges are selected sequentially from the ordered edge set to construct the minimum spanning tree. Initially, the minimum spanning tree edge set is empty, and the nodes are independent. The first edge (1,3) is selected. Nodes 1 and 3 connected by it do not belong to the same connected component and will not form a cycle. Edge (1,3) is added to the minimum spanning tree edge set. At this point, the set is {edge (1,3)}. The second edge (2,3) is selected. Nodes 2 and 3 connected by it do not belong to the same connected component (node ​​3 and node 1 are in the same component, node 2 is independent) and will not form a cycle. Edge (2,3) is added to the minimum spanning tree edge set. At this point, the set is {edge (1,3), edge (2,3)}. At this point, the minimum spanning tree edge set contains 2 edges, which is equal to the total number of discrete nodes, 3 minus one. The construction process ends. The generated minimum spanning tree connecting all discrete nodes of the target forest consists of edges (1,3) and (2,3).

[0118] Finally, the weights of all edges in the minimum spanning tree edge set are summed to generate the total length of the current habitat structure: Meters. Therefore, the total length of the current habitat structure is 8.961 meters.

[0119] S6. Obtain the total length of the baseline habitat structure of the target forest land under standard healthy succession, calculate the mathematical difference ratio between the current total habitat structure length and the baseline total habitat structure length, and generate the topological offset.

[0120] Specifically, the purpose of this step is to transform the absolute physical quantity of the current habitat structure total length calculated in step S5 into a relative and standardized evaluation index, namely, topological offset. This index can intuitively reflect the degree of deviation of the current forest vegetation spatial pattern from the ideal state, thereby providing a quantitative basis for subsequent ecological restoration decisions.

[0121] First, it is necessary to obtain the baseline habitat structure length of the target forest land under a standard healthy succession state. This baseline value represents the vegetation spatial connectivity and compactness that the forest ecosystem of this type should have in a healthy and stable state. The baseline habitat structure length can be set in various ways, for example, by analyzing and calculating data from historical healthy periods in the same area, or by simulating an ideal vegetation distribution pattern based on ecological models, considering factors such as forest land type, climate zone, and soil conditions, and then calculating it through the same process as steps S4 and S5. This value is pre-set as a reference standard in the system and is denoted as [reference value]. This value is set based on statistical analysis of long-term observation data from 50 healthy forest plots within the same type of ecological zone, ensuring its representativeness and scientific validity.

[0122] After obtaining the total length of the baseline habitat structure Then, it is compared with the total length of the current habitat structure calculated in step S5. A comparison is made. The difference between the current state and the standard state is quantified by calculating the ratio of the mathematical difference between these two values. This ratio is defined as the topological offset, denoted as . The calculation formula is as follows:

[0123] In the formula, It is the topological offset, a dimensionless pure number that represents the relative rate of change. It is the total length of the current habitat structure calculated in the current sampling period. It is the preset baseline habitat structure total length.

[0124] The physical meaning of this formula is: when Greater than hour, A positive value usually indicates that the vegetation in the woodland is becoming sparse, the average spacing between trees is increasing, resulting in an increase in the total path length required to connect all the vegetation, which may indicate vegetation degradation or death. When Less than hour, A negative value may indicate excessive vegetation density, or that the emergence of new vegetation has shortened the overall connection path, possibly indicating a stage of succession, but not necessarily a healthy state. equal hour, A value of zero indicates that the current spatial structure of the forest land is in complete agreement with the standard healthy state.

[0125] By calculating the topological offset, an absolute length value that depends on forest size and vegetation quantity is transformed into a standardized, easily comparable and interpretable percentage indicator, providing a direct quantitative basis for assessing forest health.

[0126] For example, following the example from step S5, we have already calculated the total length of the current habitat structure in the current sampling period as follows: rice.

[0127] Now, we need to obtain the total length of the baseline habitat structure for a target forest land under standard healthy succession conditions. Assume that, based on historical data and ecological models of the forest land, the total length of the baseline habitat structure is pre-defined and stored in the system as follows: Meters. This value represents the total length of the minimum spanning tree formed by the main trunks of the vegetation in this area containing three main vegetation types, under the healthiest conditions.

[0128] Next, the ratio of the mathematical difference between the current total habitat structure length and the baseline total habitat structure length is calculated to generate the topological offset. According to the formula: Calculate the difference: Meters. Calculate the ratio: .

[0129] Therefore, the generated topological offset is 0.0928, or 9.28%. This positive value indicates that the total path length required to connect all vegetation in the current forest spatial distribution pattern has increased by 9.28% compared to the standard healthy state, suggesting that the vegetation may be becoming sparser and the overall structure is degrading.

[0130] S7. Obtain basic soil sensing data of the target forest land, and perform orthogonal decoupling processing on the topological structure offset and basic soil sensing data to generate a three-dimensional physical deviation vector containing soil moisture deficit, nutrient loss equivalent and vegetation spacing increase.

[0131] In a specific embodiment of the present invention, generating a three-dimensional physical deviation vector that includes soil moisture deficit, nutrient loss equivalent, and increase in vegetation spacing includes: acquiring basic soil sensing data of the target forest land, which includes the current soil moisture value and the current soil nutrient concentration value.

[0132] The topological offset is mapped to a vegetation spatial structure degradation index, and the increase in vegetation spacing corresponding to the vegetation spatial structure degradation index is calculated.

[0133] The soil moisture deficit is calculated based on the difference between the current soil moisture value and the standard soil moisture threshold, and the nutrient loss equivalent is calculated based on the difference between the current soil nutrient concentration value and the standard soil nutrient concentration threshold.

[0134] A three-dimensional physical deviation vector is generated by orthogonally combining the soil moisture deficit, nutrient loss equivalent, and increase in vegetation spacing.

[0135] Specifically, this step aims to combine the macroscopic structural indicators (topological offset) obtained in the previous steps with field environmental factor data, performing orthogonal decoupling processing to decompose the abstract ecosystem state deviation into specific, actionable physical resource requirements. This ultimately generates a three-dimensional physical deviation vector, whose components directly correspond to the three core ecological restoration dimensions: soil moisture, soil nutrients, and vegetation density.

[0136] First, basic soil sensing data for the target forest area needs to be acquired. This data is collected in real-time or periodically by sensors deployed on-site in the forest area, mainly including current soil moisture and current soil nutrient concentration values. The current soil moisture value is denoted as... It is usually expressed as volumetric water content (percentage). The current soil nutrient concentration value is denoted as... It is usually measured in the concentration of a specific element (such as nitrogen, phosphorus, potassium) (e.g., milligrams per kilogram).

[0137] Next, the topology offset calculated in step S6 will be... This is mapped to an index of vegetation spatial structure degradation, and the increase in vegetation spacing is calculated accordingly. The topological offset is itself a dimensionless ratio that reflects the change in the total length of the minimum spanning trees. This change in length is directly related to the average spacing of vegetation in the forest. Therefore, the increase in vegetation spacing can be calculated. It is approximately correlated with the topological offset. When A positive value indicates an increase in total length, suggesting a corresponding increase in average plant spacing. (Increase in plant spacing) The calculation formula is:

[0138] In the formula, This represents the increase in plant spacing, expressed in meters. This is the average plant spacing under baseline conditions, which is based on the total length of the baseline habitat structure. and the amount of vegetation under baseline conditions The calculation method is as follows: . This is the topological offset. This formula establishes a direct mapping from macroscopic topological changes to microscopic physical distance changes.

[0139] Then, based on the acquired basic soil sensing data, the soil moisture deficit and nutrient loss equivalent are calculated. This requires comparing the current measurements with preset standard thresholds suitable for the growth of vegetation in this forest area.

[0140] Soil moisture deficit The calculation is based on the current soil moisture value. Compared with standard soil moisture threshold The difference. The standard soil moisture threshold is the lower limit of the ideal soil moisture range set according to local climate conditions and vegetation water requirements. The calculation formula is:

[0141] In the formula, It is the soil moisture deficit, measured in cubic meters, representing the amount of water that needs to be added. This is the total area of ​​the target forest land, expressed in square meters. It is the average root depth of vegetation, measured in meters, and is used to determine the soil volume for calculating water shortage. and It is a dimensionless percentage of water content by volume. This formula ensures dimensional consistency, translating differences in humidity into specific water requirements.

[0142] Nutrient loss equivalent Similarly, the calculation is based on the current soil nutrient concentration value. Compared with standard soil nutrient concentration threshold The difference. The standard soil nutrient concentration threshold is the minimum nutrient level required to maintain healthy vegetation growth. The calculation formula is:

[0143] In the formula, It is the nutrient loss equivalent, expressed in kilograms, representing the quality of nutrients that need to be replenished. It is the average bulk density of the soil, expressed in kilograms per cubic meter. and This is the nutrient concentration, expressed in milligrams per kilogram. This formula incorporates... The dimensional conversion factor converts milligrams into kilograms to ensure the consistency of physical dimensions, thereby converting concentration differences into specific nutrient requirements.

[0144] It should be noted that, in this embodiment, the typical value of the standard soil moisture threshold is 25% (volume water content), and the typical value of the standard soil nutrient concentration threshold is 80 mg / kg (based on the equivalent of fast-acting compound fertilizer). The specific values ​​are not subjectively preset, but are based on strict ecophysical benchmarks. The standard soil moisture threshold is calculated based on the field water holding capacity of a specific soil type (such as typical loam) in the target forest and the permanent wilting point of the target tree species. The lower-middle limit of the difference between the two is taken to ensure that the soil moisture is just above the critical point for plant water stress. The standard soil nutrient concentration threshold is derived by extracting historical soil background baseline survey data of the ecological zone in the healthy succession climax community stage and calibrating it in combination with the minimum physiological consumption model of the target dominant tree species during the rapid growth period. This calibration basis based on "physical field water holding capacity" and "historical ecological background line" ensures that the threshold setting is both in line with the local objective geological conditions and has certain engineering reproducibility.

[0145] Finally, the calculated soil moisture deficit will be used as a basis for further calculations. Nutrient loss equivalent Increase in plant spacing Perform orthogonal combination to generate a three-dimensional physical deviation vector. An orthogonal combination means that these three components are considered to describe the biased state of the ecosystem in three independent dimensions. This vector is represented as:

[0146] This three-dimensional physical deviation vector clearly indicates the gap between the forest land and the ideal state in terms of water, nutrients and spatial structure, providing a direct and quantifiable target for subsequent precise restoration.

[0147] For example, following the example from step S6, the topology offset is known. Assuming the target forest area is... square meters, average root depth of vegetation meters, average soil bulk density Kilograms per cubic meter. Under baseline conditions, vegetation quantity. Total length of benchmark habitat structure rice.

[0148] First, acquire basic soil sensor data: current soil moisture value. Current soil nutrient concentration values Milligrams per kilogram. The preset standard threshold is: standard soil moisture threshold. Standard soil nutrient concentration threshold Milligrams per kilogram.

[0149] Next, the increase in vegetation spacing was calculated. First, calculate the baseline average plant spacing. rice. Meters. Then, calculate the soil moisture deficit. : cubic meter.

[0150] Then calculate the nutrient loss equivalent. Note the unit conversion; the concentration difference is... mg per kilogram, i.e. kilograms per kilogram. kilogram.

[0151] Finally, the three calculation results are orthogonally combined to generate a three-dimensional physical deviation vector. : This vector indicates that in order to restore the target forest to a standard healthy state, it is necessary to replenish 35 cubic meters of water, apply 3.25 kg of nutrients, and its vegetation spatial structure will change by an average increase of 0.380 meters in plant spacing. This change may need to be corrected through measures such as replanting.

[0152] S8. Obtain the physical resource cost basis matrix configured in the management system, calculate the projection coefficient of the three-dimensional physical deviation vector on the physical resource cost basis matrix by solving the linear equation system, generate a physical material allocation list based on the projection coefficient, and send the physical material allocation list to the automated repair execution terminal for entity resource scheduling.

[0153] In a specific embodiment of the present invention, generating a physical material allocation list based on projection coefficients includes: obtaining a physical resource cost basis matrix configured in the management system, wherein the column vectors of the physical resource cost basis matrix define the standardized physical attributes and implementation equivalents of irrigation water per unit volume, compound fertilizer per unit mass, and seedlings per unit quantity.

[0154] Construct a system of linear equations with the three-dimensional physical deviation vector as the target vector and the physical resource cost basis matrix as the coefficient matrix.

[0155] The linear equations are solved by matrix inversion, and the projection coefficients of the three-dimensional physical deviation vector onto the physical resource cost basis matrix are calculated.

[0156] The projection coefficients are matched with the material types corresponding to the column vectors in the physical resource cost basis matrix to generate a physical material allocation list containing the precise scheduling quantities of irrigation water, compound fertilizer, and seedlings.

[0157] In a specific embodiment of the present invention, the physical material allocation list is sent to the automated repair execution terminal for physical resource scheduling, including: obtaining the quantity of materials to be scheduled in the physical material allocation list and the geographical coordinates of the target forest land.

[0158] The quantity and geographic coordinates of materials to be dispatched are encapsulated to generate standardized entity resource dispatch instructions.

[0159] Establish a secure communication link with the automated repair execution terminal, and send entity resource scheduling instructions to the automated repair execution terminal through the secure communication link.

[0160] Receive instruction confirmation information returned by the automated repair execution terminal, and update the resource scheduling status in the management system according to the instruction confirmation information.

[0161] Specifically, this final step transforms the ecosystem deviations diagnosed in the previous steps into a precise, automated physical resource intervention plan. The core of this process lies in using a predefined mathematical model to map abstract deviation vectors to specific quantities of resources, and ultimately dispatching automated equipment to complete the restoration task.

[0162] First, obtain a key configuration parameter from the management system: the physical resource cost basis matrix. This matrix is ​​denoted as... The physical resource cost basis matrix is ​​a 3x3 matrix that establishes a quantitative relationship between ecological restoration measures and ecological state deviations. Each column vector of the physical resource cost basis matrix corresponds to a specific restoration resource. For example, the first column defines the impact of implementing a unit volume of irrigation water (e.g., 1 cubic meter) on the ecosystem; the second column defines the impact of applying a unit mass of compound fertilizer (e.g., 1 kilogram); and the third column defines the impact of replanting a unit number of seedlings (e.g., 1 tree) on the spatial structure of the ecosystem. Each row of the matrix corresponds to one dimension of the three-dimensional physical deviation vector, namely, soil moisture deficit, nutrient loss equivalent, and increase in vegetation spacing. Therefore, the elements in the matrix... This represents the application of one unit. This resource can correct the first This matrix represents a quantity of physical deviation. Its design is based on long-term ecological experimental data and model simulations, ensuring its scientific validity and accuracy.

[0163] Next, based on the three-dimensional physical deviation vector generated in step S7 and the basis matrix of the physical resource cost of acquisition Construct a system of linear equations. In this system of equations, the three-dimensional physical deviation vector... This is the target vector, representing the deviation we hope to completely offset through resource allocation. Physical resource cost basis matrix. It is a coefficient matrix that describes the repair "efficiency" of various resources. The unknowns in the system of equations are a vector containing the quantities of the three resource scheduling operations, denoted as... The components represent the specific quantities of irrigation water, compound fertilizer, and seedlings that need to be allocated. The mathematical form of this system of linear equations is:

[0164] In the formula, It is a 3x3 physical resource cost basis matrix. It is the 3x1 resource quantity vector (i.e., projection coefficient) to be solved. It is a known 3x1 three-dimensional physical deviation vector.

[0165] To solve this system of linear equations, we need to use matrix inversion to calculate the unknown vectors. By calculating the physical resource cost basis matrix. inverse matrix Then it is compared with the three-dimensional physical deviation vector. Multiplying them yields the solution. This solution vector... Each component is a projection coefficient of the three-dimensional physical deviation vector onto the basis vectors defined by the physical resource cost basis matrix. The calculation formula is:

[0166] The solved projection coefficient vector Its three components precisely correspond to the amount of irrigation water, the quality of compound fertilizer, and the number of seedlings that need to be scheduled.

[0167] Subsequently, these calculated projection coefficients are matched with the material types corresponding to the column vectors in the physical resource cost basis matrix to generate a clear physical resource allocation list. This list details the precise quantities of the three resources required to restore the target forest.

[0168] Finally, the entire process enters the automated execution phase. The system obtains the quantity of materials to be dispatched from the physical material allocation list and the geographical coordinates of the target forest land. This information is encapsulated into a standardized physical resource dispatch instruction. The system establishes a secure communication link with the automated repair execution terminals deployed on-site (such as intelligent irrigation systems, drone fertilization platforms, etc.) and sends the instruction through this link. After receiving the instruction, the terminal executes the corresponding irrigation, fertilization, or replanting tasks and returns instruction confirmation information to the management system. The management system updates the resource dispatch status based on the confirmation information, completing the entire closed-loop automated repair process.

[0169] For example, following the example from step S7, we have already generated the three-dimensional physical deviation vector. The units of measurement are cubic meters, kilograms, and meters, respectively.

[0170] First, obtain the configured physical resource cost basis matrix from the management system. Assume that the matrix is ​​set according to local ecological conditions as follows: The meaning of this matrix is ​​as follows: The first column represents irrigation water; 1 cubic meter of water can replenish a 1 cubic meter water shortage, with no effect on nutrients or plant spacing. The second column represents compound fertilizer; 1 kilogram of compound fertilizer can replenish a 1 kilogram of nutrient loss equivalent, with no effect on water or plant spacing. The third column represents seedlings; replanting one seedling, based on the average plant spacing of the forest baseline, can correct the increase in vegetation spacing by 4.100 meters, with no effect on water or nutrients.

[0171] Next, construct a system of linear equations. : The projection coefficients are solved by matrix inversion. .matrix inverse matrix for: .

[0172] calculate : Since the number of saplings must be an integer, the system will... The treatment is set to 1, meaning that at least one tree should be replanted if there is any structural deviation.

[0173] Therefore, the generated physical material allocation list is as follows: irrigation water: 35 cubic meters, compound fertilizer: 3.25 kilograms, sapling: 1 tree.

[0174] Finally, obtain the geographical coordinates of the target forest, such as (116.3 degrees east longitude, 39.9 degrees north latitude). Encapsulate the material scheduling quantity (35, 3.25, 1) and geographical coordinates to generate standardized entity resource scheduling instructions, and send them to the automated repair execution terminal through a secure communication link.

[0175] Reference Figure 2 The second aspect of the present invention provides a system for calculating the trajectory offset of forest degradation habitat succession, comprising: a static physical anchor point set generation module, a spatial affine transformation parameter generation module, an unvariable three-dimensional point cloud data generation module, an undirected complete graph construction module, a current habitat structure total length generation module, a topological structure offset generation module, a three-dimensional physical deviation vector generation module, and a material list mapping and scheduling module.

[0176] The static physical anchor point set generation module is connected to the spatial affine transformation parameter generation module. The spatial affine transformation parameter generation module is connected to the unchanging 3D point cloud data generation module. The unchanging 3D point cloud data generation module is connected to the undirected complete graph construction module. The undirected complete graph construction module is connected to the current habitat structure total length generation module. The current habitat structure total length generation module is connected to the topology offset generation module. The topology offset generation module is connected to the 3D physical deviation vector generation module. The 3D physical deviation vector generation module is connected to the material list mapping and scheduling module.

[0177] The static physical anchor point set generation module acquires the three-dimensional laser point cloud physical data of the target forest land under different sampling periods, performs morphological elevation filtering, extracts the three-dimensional physical coordinate point set of vegetation base that is close to the ground and unaffected by wind disturbance, and generates the static physical anchor point set.

[0178] The spatial affine transformation parameter generation module uses the set of stationary physical anchor points in the first sampling period as the absolute reference and employs the singular value decomposition algorithm to calculate the rotation and translation matrices of the set of stationary physical anchor points in any subsequent sampling period relative to the absolute reference, thereby generating spatial affine transformation parameters.

[0179] The non-deformable 3D point cloud data generation module uses spatial affine transformation parameters to perform an inverse rigid body transformation on the physical data of the 3D laser point cloud in any subsequent sampling period, generating non-deformable 3D point cloud data with strictly aligned physical coordinates.

[0180] The undirected complete graph construction module extracts the centroid 3D physical coordinates of all vegetation trunks from the unvariable 3D point cloud data, uses the centroid 3D physical coordinates as discrete nodes, and uses the Euclidean physical space distance between discrete nodes as the edge length weight to construct an undirected complete graph.

[0181] The current habitat structure total length generation module runs the minimum spanning tree algorithm on the undirected complete graph, calculates the minimum spanning tree connecting all discrete nodes of the target forest, and sums up the topological edge lengths of the minimum spanning tree to generate the current habitat structure total length.

[0182] The topology offset generation module obtains the total length of the baseline habitat structure when the target forest is in a standard healthy succession state, calculates the mathematical difference ratio between the current total habitat structure length and the baseline total habitat structure length, and generates the topology offset.

[0183] The three-dimensional physical deviation vector generation module acquires the basic soil sensing data of the target forest land, performs orthogonal decoupling processing on the topological structure offset and the basic soil sensing data, and generates a three-dimensional physical deviation vector containing soil moisture deficit, nutrient loss equivalent, and vegetation spacing increase.

[0184] The material list mapping and scheduling module obtains the physical resource cost basis matrix configured in the management system, calculates the projection coefficient of the three-dimensional physical deviation vector on the physical resource cost basis matrix by solving a system of linear equations, generates a physical material allocation list based on the projection coefficient, and sends the physical material allocation list to the automated repair execution terminal for entity resource scheduling.

[0185] The above content is merely an example and illustration of the concept of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described, or use similar methods to replace them, as long as they do not deviate from the concept of the invention or exceed the scope defined by the present invention, and all such modifications and additions should fall within the protection scope of the present invention.

Claims

1. A method for calculating the offset of forest degradation habitat succession trajectory, characterized in that, include: S1. Obtain the three-dimensional laser point cloud physical data of the target forest land under different sampling periods, and perform morphological elevation filtering to extract the three-dimensional physical coordinate point set of the vegetation base that is close to the ground and unaffected by wind disturbance, and generate a set of static physical anchor points. S2. Using the set of stationary physical anchor points in the first sampling period as the absolute reference, the singular value decomposition algorithm is used to calculate the rotation matrix and translation matrix of the set of stationary physical anchor points in any subsequent sampling period relative to the absolute reference, and spatial affine transformation parameters are generated. S3. Using spatial affine transformation parameters, perform inverse rigid body transformation on the physical data of the three-dimensional laser point cloud in any subsequent sampling period to generate non-deformable three-dimensional point cloud data with strictly aligned physical coordinates. S4. Extract the centroid three-dimensional physical coordinates of all vegetation trunks from the non-deformable three-dimensional point cloud data, use the centroid three-dimensional physical coordinates as discrete nodes, and use the Euclidean physical space distance between discrete nodes as the edge length weight to construct an undirected complete graph. S5. Run the minimum spanning tree algorithm on the undirected complete graph to calculate the minimum spanning tree connecting all discrete nodes of the target forest, and sum the topological edge lengths of the minimum spanning tree to generate the total length of the current habitat structure. S6. Obtain the total length of the baseline habitat structure of the target forest land under the standard healthy succession state, calculate the mathematical difference ratio between the current total habitat structure length and the baseline total habitat structure length, and generate the topological structure offset. S7. Obtain basic soil sensing data of the target forest land, perform orthogonal decoupling processing on the topological structure offset and basic soil sensing data, and generate a three-dimensional physical deviation vector containing soil moisture deficit, nutrient loss equivalent and vegetation spacing increase. S8. Obtain the physical resource cost basis matrix configured in the management system, calculate the projection coefficient of the three-dimensional physical deviation vector on the physical resource cost basis matrix by solving the linear equation system, generate a physical material allocation list based on the projection coefficient, and send the physical material allocation list to the automated repair execution terminal for entity resource scheduling.

2. The method for calculating the offset of forestry degraded habitat succession trajectory according to claim 1, characterized in that, The generation of the set of static physical anchor points includes: Acquire three-dimensional laser point cloud physical data of the target forest land under different sampling periods; Morphological elevation filtering is performed on the 3D laser point cloud physical data to separate the ground point cloud and non-ground point cloud. Point cloud data with elevation values ​​within the near-ground threshold range are extracted from non-ground point clouds and used as a set of three-dimensional physical coordinate points for the vegetation base. Density clustering is used to denoise the three-dimensional physical coordinate point set at the base of vegetation, eliminating isolated noise points and generating a set of static physical anchor points.

3. The method for calculating the offset of forestry degraded habitat succession trajectory according to claim 1, characterized in that, The generated spatial affine transformation parameters include: Extract the set of stationary physical anchor points in the first sampling period as the absolute reference, and extract the set of stationary physical anchor points in any subsequent sampling period as the set of registration points; Calculate the centroid coordinates of the absolute datum and the centroid coordinates of the point set to be registered, and then perform centroid removal processing on the absolute datum and the point set to be registered respectively to generate the datum centroid removal point set and the point set to be registered centroid removal point set; Construct the covariance matrix between the baseline decentroid set and the decentroid set to be registered, and process the covariance matrix using the singular value decomposition algorithm to generate the rotation matrix; The translation matrix is ​​calculated based on the rotation matrix, the centroid coordinates of the absolute reference, and the centroid coordinates of the set of points to be registered. The rotation matrix and the translation matrix are then combined to generate spatial affine transformation parameters.

4. The method for calculating the offset of forestry degraded habitat succession trajectory according to claim 1, characterized in that, The generated, physically coordinate-aligned, non-deformable 3D point cloud data includes: Obtain each original 3D data point from the 3D laser point cloud physical data of any subsequent sampling period; Perform matrix multiplication on each original 3D data point and the rotation matrix in the spatial affine transformation parameters to generate rotation-aligned data points; The rotation-aligned data points are generated by performing a vector addition operation with the translation matrix in the spatial affine transformation parameters; Perform matrix multiplication and vector addition operations on all original 3D data points in the 3D laser point cloud physical data of any subsequent sampling period to generate deformation-free 3D point cloud data with strictly aligned physical coordinates.

5. The method for calculating the offset of forestry degraded habitat succession trajectory according to claim 1, characterized in that, The construction of the undirected complete graph includes: Cylindrical model fitting was performed on the non-deformable 3D point cloud data to identify the vegetation trunk point cloud set in the non-deformable 3D point cloud data. Calculate the geometric center coordinates of the vegetation trunk point cloud set in three-dimensional space, and use the geometric center coordinates as the three-dimensional physical coordinates of the centroid of the vegetation trunk. Define the three-dimensional physical coordinates of the centroid as discrete nodes in the graph theory model, and calculate the Euclidean physical space distance between any two discrete nodes; Using Euclidean physical distance as the edge length weight connecting two corresponding discrete nodes, we connect all discrete nodes to construct an undirected complete graph.

6. The method for calculating the offset of forestry degraded habitat succession trajectory according to claim 1, characterized in that, The generation of the total length of the current habitat structure includes: Arrange all edges in the undirected complete graph in ascending order of edge length weight to generate an ordered set of edges; Select the edge with the smallest weight from the ordered edge set one by one, and determine whether the selected edge forms a cycle in the selected edge set. If no loop is formed, the selected edge is added to the minimum spanning tree edge set until the number of edges in the minimum spanning tree edge set is equal to the total number of discrete nodes minus one, thus generating a minimum spanning tree connecting all discrete nodes of the target forest. The weights of the edge lengths of all edges in the minimum spanning tree edge set are summed to generate the total length of the current habitat structure.

7. The method for calculating the offset of forestry degraded habitat succession trajectory according to claim 1, characterized in that, The generation of the three-dimensional physical deviation vector, which includes soil moisture deficit, nutrient loss equivalent, and increase in vegetation spacing, includes: Acquire basic soil sensing data of the target forest land, including current soil moisture and current soil nutrient concentration; The topological offset is mapped to the vegetation spatial structure degradation index, and the increase in vegetation spacing corresponding to the vegetation spatial structure degradation index is calculated. The soil moisture deficit is calculated based on the difference between the current soil moisture value and the standard soil moisture threshold, and the nutrient loss equivalent is calculated based on the difference between the current soil nutrient concentration value and the standard soil nutrient concentration threshold. A three-dimensional physical deviation vector is generated by orthogonally combining the soil moisture deficit, nutrient loss equivalent, and increase in vegetation spacing.

8. The method for calculating the offset of forestry degraded habitat succession trajectory according to claim 1, characterized in that, The process of generating a physical material allocation list based on the projection coefficient includes: Obtain the physical resource cost basis matrix configured in the management system. The column vectors of the physical resource cost basis matrix define the standardized physical attributes and implementation equivalents of irrigation water per unit volume, compound fertilizer per unit mass, and seedlings per unit quantity. Construct a system of linear equations with the three-dimensional physical deviation vector as the target vector and the physical resource cost basis matrix as the coefficient matrix; Solve the linear equation system by matrix inversion, and calculate the projection coefficient of the three-dimensional physical deviation vector onto the physical resource cost basis matrix. The projection coefficients are matched with the material types corresponding to the column vectors in the physical resource cost basis matrix to generate a physical material allocation list containing the precise scheduling quantities of irrigation water, compound fertilizer, and seedlings.

9. The method for calculating the offset of forestry degraded habitat succession trajectory according to claim 1, characterized in that, The step of sending the physical material allocation list to the automated repair execution terminal for physical resource scheduling includes: Obtain the quantity of materials to be dispatched from the physical material allocation list and the geographical coordinates of the target forest area; The quantity and geographic coordinates of materials to be dispatched are encapsulated to generate standardized entity resource dispatch instructions; Establish a secure communication link with the automated repair execution terminal, and send entity resource scheduling instructions to the automated repair execution terminal through the secure communication link; Receive instruction confirmation information returned by the automated repair execution terminal, and update the resource scheduling status in the management system according to the instruction confirmation information.

10. A system for calculating the offset of forestry degraded habitat succession trajectory, characterized in that, include: The static physical anchor point set generation module acquires the three-dimensional laser point cloud physical data of the target forest land under different sampling periods, performs morphological elevation filtering, extracts the three-dimensional physical coordinate point set of vegetation base that is close to the ground and unaffected by wind disturbance, and generates the static physical anchor point set. The spatial affine transformation parameter generation module uses the set of stationary physical anchor points in the first sampling period as the absolute reference and employs the singular value decomposition algorithm to calculate the rotation and translation matrices of the set of stationary physical anchor points in any subsequent sampling period relative to the absolute reference, thereby generating spatial affine transformation parameters. The non-deformable 3D point cloud data generation module uses spatial affine transformation parameters to perform inverse rigid body transformation on the physical data of 3D laser point cloud in any subsequent sampling period to generate non-deformable 3D point cloud data with strictly aligned physical coordinates. The undirected complete graph construction module extracts the centroid 3D physical coordinates of all vegetation trunks from the unvariable 3D point cloud data, uses the centroid 3D physical coordinates as discrete nodes, and uses the Euclidean physical space distance between discrete nodes as the edge length weight to construct an undirected complete graph. The current habitat structure total length generation module runs the minimum spanning tree algorithm on the undirected complete graph, calculates the minimum spanning tree connecting all discrete nodes of the target forest, and sums up all topological edge lengths of the minimum spanning tree to generate the current habitat structure total length. The topology offset generation module obtains the total length of the baseline habitat structure of the target forest under standard healthy succession, calculates the mathematical difference ratio between the current total habitat structure length and the baseline total habitat structure length, and generates the topology offset. The three-dimensional physical deviation vector generation module acquires the basic soil sensing data of the target forest land, performs orthogonal decoupling processing on the topological structure offset and the basic soil sensing data, and generates a three-dimensional physical deviation vector containing soil moisture deficit, nutrient loss equivalent and vegetation spacing increase. The material list mapping and scheduling module obtains the physical resource cost basis matrix configured in the management system, calculates the projection coefficient of the three-dimensional physical deviation vector on the physical resource cost basis matrix by solving a system of linear equations, generates a physical material allocation list based on the projection coefficient, and sends the physical material allocation list to the automated repair execution terminal for entity resource scheduling.

Images