A modeling method, system, device and medium for the whole process of black locust degradation based on L-system

Abstract

The present invention discloses a method, system, device and medium for modeling the whole process of black locust degradation based on the L-system, which relates to the technical field of biological growth forms. The method includes: obtaining the basic data of black locust from field measurement data and literature materials, and extracting L-system rules based on the basic data; the L-system rules include main trunk rules, branch pattern rules and leaf simulation rules; performing partial modeling of black locust trees according to the L-system rules to obtain main trunk and branch models, leaf models and root models respectively; combining single black locust tree models according to the main trunk and branch models, the leaf models and the root models, and simulating the dead tip rate of black locust trees to obtain a change model of the whole process of black locust degradation. The present invention can improve the accuracy and applicability of the model.

Description

Technical Field

[0001] The present invention relates to the technical field of biological growth morphology, and particularly to a modeling method, system, device and medium for the whole process of Robinia pseudoacacia degradation based on the L-system. Background Art

[0002] Single tree modeling, as an important technical means in forest resource monitoring and ecological research, has been widely used in recent years in fields such as precision forest management, ecosystem service assessment, and biodiversity conservation. Single tree modeling provides high-precision data information for the dynamic monitoring and precision management of forest resources through three-dimensional modeling of single trees, position recognition, attribute estimation, and growth prediction.

[0003] Currently, the mainstream methods of single tree modeling can be divided into the following categories: 1) Rule-based modeling efficiently generates complex branching structures through rule recursion and is widely used in virtual scenarios; 2) Functional-structural models combine the physiological processes and morphological structures of plants and are suitable for ecological research. 3) Geometric branching models are commonly used in film and games and are easy to control the tree morphology. 4) Data-driven models use LiDAR and photogrammetry data to generate realistic tree morphologies and are commonly used in forest monitoring. 5) Hybrid modeling has gradually become a trend, combining the advantages of rule-based, physical, and data-driven models, taking into account the efficiency, realism, and multi-scenario adaptability of modeling, making it applicable to complex scenarios such as ecology and virtual reality. These methods have achieved a balance among realism, computational efficiency, and flexibility, providing feasible solutions for tree modeling in different fields.

[0004] The L-system (Lindenmayer system) is a rule-based modeling method mainly composed of a symbol alphabet, a production rule set, an initial axiom string, and a translation mechanism. The symbol alphabet contains basic building blocks, while the production rules define how each symbol generates longer symbol strings through replacement or expansion. The initial axiom string serves as the starting point of the model, and new strings can be generated by repeatedly applying the production rules. The generated strings are transformed into geometric structures through the translation mechanism, enabling the L-system to not only describe the growth morphology of plants but also simulate other biological morphologies and generate self-similar fractal structures. This feature makes the L-system an ideal tool for complex ecological modeling and growth simulation. By combining mathematical models with biological characteristics, single tree modeling based on the L-system plays various important roles in practical production such as ecological restoration and landscape design, environmental impact assessment, biotechnology research, and forestry resource management.

[0005] Although the traditional single-tree modeling method based on the L-system can simulate the growth form of trees, it has problems such as complex parameter settings, static models, insufficient natural realism, large computational volume, and poor generality. Moreover, it lacks dynamic simulation of environmental factors, tree physiological processes, and ecological relationships, so it is limited in practical applications. Summary of the Invention

[0006] The purpose of the present invention is to provide a modeling method, system, device, and medium for the whole process of black locust degradation based on the L-system, which can improve the model accuracy and applicability.

[0007] To achieve the above purpose, the present invention provides the following solutions:

[0008] A modeling method for the whole process of black locust degradation based on the L-system includes:

[0009] Obtain the basic data of black locust; the basic data is determined from field measurement data and literature materials;

[0010] Extract the L-system rules based on the basic data; the L-system rules include the main trunk rules, branch pattern rules, and leaf simulation rules;

[0011] Conduct partial modeling of black locust trees according to the L-system rules to obtain the main trunk and branch models, leaf models, and root models respectively;

[0012] Combine the single black locust tree model according to the main trunk and branch models, the leaf model, and the root model, and simulate the dieback rate of black locust trees to obtain the change model of the whole process of black locust degradation.

[0013] Optionally, the extraction formula of the L-system rules includes:

[0014] The definition of the main trunk rule is the part below the live crown, which is expressed by the formula @!(w)F(l), where @ means the following character is an axiom character,!(w) means the radius of the main trunk drawn is w, and F(l) means drawing a frustum with a length of l as the branch internode;

[0015] The branch pattern rule is expressed by the formula p1: A(l,w) →!(w)F(l)[&(a0)B(l×r2,w×wr)][ / (a1)B(l×r2,w×wr)][&(a0)B(l×r2,w×wr)][&(a1)B(l×r2,w×wr)], indicating that the vertex of the main axis A generates an upward internode F and four side branch vertices B in each step of derivation; where, l×r2 represents the length of the branch, w×wr represents the thickness of the branch, and &(a0), / (a1), &(a1) are all adjustment parameters of the branch morphological structure;

[0016] The leaf simulation rule is expressed by the formula L(a, b, c), where a represents the length of the branch segment, b represents the number of leaves on the branch segment, and c represents the size of the leaf.

[0017] Optionally, the modeling process of the main trunk and branch model includes:

[0018] Derive using the production rule A(l, w). The vertex of the main axis A generates a node F and four lateral branches B in each step of the derivation, completing one round of branch growth. Then, based on the iteration step, gravity, and random length, perform multiple rounds of simulation of the natural growth structure of the branches to obtain the main trunk and branch model.

[0019] Optionally, the modeling process of the leaf model includes:

[0020] Use 3ds Max modeling software to perform single-leaf surface modeling, and determine the coordinates of the skeleton points and the network construction index between the points.

[0021] Based on the coordinates of the skeleton points and the network construction index between the points, perform trigonometric function modeling to obtain the Nurbs surface of the leaf shape.

[0022] Convert the Nurbs surface of the leaf shape into a mesh, and perform quadrilateral mesh reconstruction. Adjust the growth shape of the leaf within the mesh to obtain the leaf model.

[0023] Optionally, the process of converting the Nurbs surface of the leaf shape into a mesh specifically includes:

[0024] 1) Map the points on the Nurbs surface to three-dimensional space and represent them as: z = f(x, y);

[0025] 2) Perform mesh division, and set the sampling area range as [x min , x max and [y min , y max ;

[0026] 3) Determine the mesh resolution, including the number of sampling points in the x direction and the y direction, which are N x and N y respectively. Calculate the sampling step size in each direction as:

[0027]

[0028] 4) Generate all sampling points in the two-dimensional mesh by sampling the x value and the y value, and represent them as:

[0029] x i = x min + i·Δx where i = 0, 1, ……, Nx -1;

[0030] y i = y min + j·Δy where j = 0, 1, ……, N y -1;

[0031] 5) For the i-th sampling point (x i , y j ), generate the corresponding z value: z ij = f(x i , y j ), and combine the x value and the y value to obtain the three-dimensional coordinates (x i , y j , z ij );

[0032] 6) Connect each sampling point to form a network. In a two-dimensional grid, the sampling points form a grid surface by forming quadrilaterals, and divide each small unit of each grid surface into two triangles:

[0033] The first triangle: (x i , y j , z ij ), (x i+1 , y j , z i+1,j ), (x i , y i+1 , z ij+1 );

[0034] The second triangle: (x i+1 , y j , z i+1,j ), (x i+1 , y i+1 , z i+1,j+1 ), (x i , y j+1 , z ij+1 );

[0035] 7) Since photosynthesis is considered when modeling the blade, calculate the normal vector of each grid surface. The formula is as follows:

[0036] n = P1 × P2;

[0037] In the formula, P1 and P2 are edge vectors derived from adjacent vertices. Calculate the normal vector using the vertex (x i , y j , z ij ) and adjacent vertices to obtain:

[0038]

[0039] Set corresponding rules to align the seam points, set the minimum and maximum numbers of the starting quadrilateral meshes, and set corresponding rules for the lengths of the largest and smallest edges according to the preset aspect ratio to complete the mesh conversion.

[0040] Optionally, the modeling process of the root system model includes:

[0041] Use the A* algorithm to determine the shortest path from the root system to the target water source or nutrient source, perform feedback adjustment according to the environmental attributes affecting the growth of the root system, and perform dynamic growth using the slime mold algorithm to obtain the root system model.

[0042] Optionally, the simulation process of the withered tip rate of Robinia pseudoacacia includes height filtering, random ratio filtering, and spindle shape filtering; wherein, the height filtering is to remove the leaves exceeding the set height and retain the leaves below the set height; the random ratio filtering is to delete the leaves according to the set ratio; the spindle shape filtering is to model the spindle shape of the set screening rules, use the generated spindle shape model as the outer frame boundary of the single Robinia pseudoacacia model, and remove the leaves with the center point outside the boundary.

[0043] The present invention also provides a modeling system for the whole process of Robinia pseudoacacia degradation based on the L-system, including:

[0044] A data acquisition unit for obtaining the basic data of Robinia pseudoacacia; the basic data is determined from field measured data and literature materials;

[0045] A rule extraction unit for extracting L-system rules based on the basic data; the L-system rules include main trunk rules, branch pattern rules, and leaf simulation rules;

[0046] A Robinia pseudoacacia distribution modeling unit for performing Robinia pseudoacacia distribution modeling according to the L-system rules to obtain a main trunk and branch model, a leaf model, and a root system model respectively;

[0047] A Robinia pseudoacacia withered tip rate simulation unit for combining a single Robinia pseudoacacia model according to the main trunk and branch model, the leaf model, and the root system model, and simulating the withered tip rate of Robinia pseudoacacia to obtain a change model of the whole process of Robinia pseudoacacia degradation.

[0048] The present invention also provides an electronic device, including a memory and a processor, the memory is used to store a computer program, and the processor runs the computer program to make the electronic device execute the method for modeling the whole process of Robinia pseudoacacia degradation based on the L-system as described above.

[0049] The present invention also provides a computer-readable storage medium, which stores a computer program, and when the computer program is executed by a processor, it implements the method for modeling the whole process of Robinia pseudoacacia degradation based on the L-system as described above.

[0050] According to the specific embodiments provided by the present invention, the following technical effects are disclosed by the present invention:

[0051] The present invention discloses a method, a system, a device and a medium for modeling the whole process of Robinia pseudoacacia degradation based on the L-system. The method includes obtaining the basic data of Robinia pseudoacacia from field measured data and literature materials, and extracting the L-system rules based on the basic data; the L-system rules include a main trunk rule, a branch pattern rule and a leaf simulation rule; according to the L-system rules, a Robinia pseudoacacia distribution model is established, and a main trunk and branch model, a leaf model and a root system model are obtained respectively; according to the main trunk and branch model, the leaf model and the root system model, a single Robinia pseudoacacia model is combined, and the withered tip rate of Robinia pseudoacacia is simulated to obtain a change model of the whole process of Robinia pseudoacacia degradation. The present invention can improve the accuracy and applicability of the model. Brief Description of the Drawings

[0052] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required to be used in the embodiments. Obviously, the following described drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0053] Figure 1 is a schematic flow chart of the method for modeling the whole process of Robinia pseudoacacia degradation based on the L-system of the present invention;

[0054] Figure 2 is a simulation diagram of the branch model in this embodiment;

[0055] Figure 3 is a simplified diagram of the branch in this embodiment;

[0056] Figure 4 is a schematic diagram of the secondary and tertiary branch models in this embodiment;

[0057] Figure 5 is a schematic diagram of the leaf skeleton in this embodiment;

[0058] Figure 6 is a schematic diagram of the rendered single leaf model in this embodiment;

[0059] Figure 7 is a schematic diagram of the horizontal leaf bending function in this embodiment;

[0060] Figure 8 is a schematic diagram of the vertical leaf bending function in this embodiment;

[0061] Figure 9 is a schematic diagram of the leaf surface in this embodiment;

[0062] Figure 10 Schematic diagram of blade grid reconstruction in this embodiment;

[0063] Figure 11 Schematic diagram of simulating a real tip blade in this embodiment;

[0064] Figure 12 Schematic diagrams before and after random adjustment of the blade in this embodiment;

[0065] Figure 13 Schematic diagram of not meeting the simulated natural state in this embodiment;

[0066] Figure 14 Schematic diagram of the maximum form of the sun-facing area in this embodiment;

[0067] Figure 15 Schematic diagram of adjusting the blade size in this embodiment;

[0068] Figure 16 Model diagram of Robinia pseudoacacia under different parameters in this embodiment;

[0069] Figure 17 Model diagram of Robinia pseudoacacia after height filtering in this embodiment;

[0070] Figure 18 Model diagram of Robinia pseudoacacia with a withered tip rate of 30% and leaves above 7 meters removed in this embodiment;

[0071] Figure 19 Model diagram of Robinia pseudoacacia with a withered tip rate of 30%, leaves above 7 meters removed, and simulating the spindle shape in this embodiment;

[0072] Figure 20 Schematic diagram of the 3D model structure of the root system in this embodiment;

[0073] Figure 21 Schematic diagram of the morphology imported into the analysis model software in this embodiment. Detailed implementation manners

[0074] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0075] The purpose of the present invention is to provide a modeling method, system, device, and medium for the whole process of Robinia pseudoacacia degradation based on the L-system, which can improve the model accuracy and applicability.

[0076] To make the above objects, features, and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0077] In the prior art, the following problems generally exist: (1) It is insufficient in representing biodiversity, and many natural trees are difficult to fully simulate through simple L-system rules, especially some special growth patterns, tree species characteristics, or ecological adaptabilities. (2) The L-system is usually based on an idealized growth model and insufficiently considers environmental impacts (such as gravity, wind, light, etc.) during the actual growth process, which may result in the generated tree morphology being biologically less realistic. (3) The models generated by the L-system mainly focus on the basic structure of the tree and are insufficient in representing details (such as the shape of leaves). To achieve a high level of detail, additional model refinement steps are usually required, thereby increasing the complexity and computational amount of modeling. (4) In the modeling of plant roots based on the L-system, the current relatively mature methods are for the root modeling of crops such as rice and corn. Since the roots of trees and crops have obvious differences in structure, function, ecological adaptability, etc., other algorithms need to be combined to simulate the growth of roots. (5) In the traditional application of the L-system, the ecological phenomenon of the dead shoot rate is often not taken into account, and the generated trees show an ideal lush state, lacking a realistic representation of the withered parts. (6) Currently, the research on a single tree species is far from meeting the needs of multi-tree species modeling. When the tree species are different, the technical methods used are not applicable, and the current research fails to construct tree models with any degree of degradation.

[0078] Therefore, the present invention provides a modeling method for the entire process of Robinia pseudoacacia degradation as Figure 1 shown, including:

[0079] Step 100: Obtain the basic data of Robinia pseudoacacia; the basic data is determined from field measurement data and literature.

[0080] Step 200: Extract L-system rules based on the basic data; the L-system rules include main trunk rules, branch pattern rules, and leaf simulation rules.

[0081] Step 300: Perform partial modeling of Robinia pseudoacacia trees according to the L-system rules to obtain main trunk and branch models, leaf models, and root models respectively.

[0082] Step 400: Combine single Robinia pseudoacacia tree models according to the main trunk and branch models, the leaf models, and the root models, and simulate the dead shoot rate of Robinia pseudoacacia trees to obtain a change model for the entire process of Robinia pseudoacacia degradation.

[0083] As a specific embodiment, the following embodiments are provided to detail the processing procedures of the above steps.

[0084] First, the basic concept of the L-system is described: The L-system is a parallel rewriting system consisting of a symbol alphabet, a set of production rules, an initial axiom string, and a translation mechanism. The symbol alphabet contains the basic building blocks of the system, and the production rules specify how each symbol generates a larger symbol string through substitution or expansion. The initial axiom string serves as the starting point, and new strings are generated by iteratively applying the production rules. The translation mechanism converts the generated strings into geometric structures, enabling the L-system to describe the development process of biological structures, simulate various biological forms, and be used to generate self-similar fractal structures.

[0085] After presenting the simulation language, this embodiment focuses on the in-depth simulation of the morphological structure of plants. This embodiment considers using the L-system as a basis to describe a locust tree with components such as roots, trunks, and side branches. In this context, the following definitions are introduced: A root node is a node with special properties. Edges (line segments) extend from the root node to the termination node, forming a path. Inner nodes are in the path and are connected to at least one successor edge. The tip is the termination point, marking the end of the tree, which is a node without a successor edge. Straight branches and side branches form a sequence of line segments from the bottom to the top of the tree, corresponding to the trunk and branches of the tree. These trunks and branches can be divided into levels 0, 1, 2, etc., forming a representation of an axial tree.

[0086] The significant characteristics of a normally growing locust tree can be obtained from field measured data and literature:

[0087] (1) Crown shape characteristics: The crown of a healthy locust tree shows a monopodial tree form with a significant apical dominance, and the trunk is straight and solid. Its crown shape exhibits obvious changes at different growth stages. It is round and harmonious in the young age, approximately a rotational paraboloid in the middle age, and approximately a frustum shape in the mature age. This monopodial tree form and the changes in the crown shape at different stages reflect the morphological adjustment and adaptability of the locust tree during its growth and development.

[0088] (2) Branch characteristics: The branch structure of the locust tree is relatively complex, showing whorled branches of 4 - 5, spreading radially. The number of branch disks formed annually is closely related to the height growth level. During the peak height growth period, an average of 5 - 6 branch disks can occur. As the tree age increases, the branching rate shows a gradually increasing trend. The branching angle of the first-order branches is mainly distributed between 60° and 90°, while the branching degree of the second-order branches is mostly distributed between 30° and 90°. In addition, the azimuth angle of the first-order branches is evenly distributed, while the azimuth angle of the second-order branches is mainly distributed in the horizontal direction.

[0089] (3) Leaf characteristics: The leaves of Robinia pseudoacacia are densely arranged in a spiral and alternate pattern, growing in a drooping manner. On the lateral branches, the leaf bases are twisted to form a two-row arrangement. The leaf shape is linear-lanceolate, with serrated edges. The leaves on vegetative branches are relatively larger, about 3 - 6 cm in length and about 0.4 - 0.5 cm in width, while the leaves on reproductive branches are relatively smaller, less than 2 cm in length. This arrangement and morphology of the leaves provide the basis for photosynthesis and reproductive activities of Robinia pseudoacacia, and also add uniqueness to its appearance characteristics.

[0090] Based on this, first, L-system is used for rule extraction. The L-system rule extraction is based on the field measured data of young Robinia pseudoacacia trees and the descriptions of the morphological structure characteristics of Robinia pseudoacacia in relevant literature. In this embodiment, the constraint rules for L-system modeling are extracted. First, focus on the rules of the main trunk. The main trunk is defined as the part below the lowest branch and is expressed in the rule axiom. For example: "@!(0.1)F(0.4)", where @ indicates that the following characters are axiom characters,!(0.1) represents the radius of the main trunk to be drawn, and F(0.4) represents drawing a frustum with a length of 0.4 as the internode of the branch. These parameters are adjusted in the rules according to the measured data of the height below the lowest branch of the Robinia pseudoacacia tree.

[0091] Secondly, the branching pattern is another key feature of the branching structure of Robinia pseudoacacia trees, usually showing a whorled-branch pattern. This branching pattern is represented by L-system rules, such as "p1:A(l,w)→!(w)F(l)[&(a0)B(l×r2,w×wr)][ / (a1)B(l×r2,w×wr)][&(a0)B(l×r2,w×wr)][&(a1)B(l×r2,w×wr)]", which means that the vertex of the main axis A generates an upward main trunk internode F and four lateral branch vertices B in each step of derivation.

[0092] Furthermore, the branch length and thickness are controlled by the parameters in the characters. For example, in the above production p1, the first and second parameters in B(l×r2,w×wr) represent the length and thickness of the branch respectively. The inclination angle and azimuth angle of the branch are also controlled by parameters. For example, the parameters in &(a0), / (a1), &(a1) in the above production are used to adjust the morphological structure of the branch.

[0093] Finally, regarding the simulation of leaves, in this embodiment, it is considered that the leaves of Robinia pseudoacacia trees are of a specific type and are distributed in a spiral pattern on the surface of the branches. Taking a branch segment as the basic unit, the uniform distribution of leaves is realized through L-system rules, the leaf azimuth angle is randomly distributed, and the leaf inclination angle varies within a certain range. The normalized leaf size is set in the leaf point file, and the specific rule is shown as "L(a,b,c)", where a represents the length of the branch segment, b represents the number of leaves on this segment of the branch, and c represents the size of the leaf. This rule takes into account the actual growth characteristics of Robinia pseudoacacia trees to more realistically simulate its appearance.

[0094] Furthermore, based on the above rules, a hierarchical model of the locust tree is constructed, including the modeling of the main trunk and branches, leaf modeling, and root system modeling.

[0095] As the first part, the modeling of the main trunk and branches is carried out:

[0096] Through the derivation of production rule A(I, w), the vertex of the main axis A generates an internode F and four lateral branches B in each step of the derivation. These four lateral branches simulate the growth of a whorl of branches of the locust tree, where production rule B(I, w) indicates that the vertex of the lateral axis B generates an internode F and two lateral branches C. In the same whorl, the introduction of a section of internode between two lateral branches aims to simulate the non-strict whorled arrangement of lateral branches, that is, the attachment points of the whorled branches in the same whorl are not at the same height on the main trunk. Each lateral branch C represents an internode F. When the iteration step size is 16, the simulated branch model and the simplified diagram are as shown in Figure 2 and Figure 3 shown.

[0097] On this basis, the influence of gravity is added, and the branches show a curved state. The degree of curvature is determined by different lengths and different numbers of secondary branches. It is obtained from the given formula. In this simulation example, the method of manually inputting the quantity is adopted to improve the running speed of the computer. Due to the influence of photosynthesis, the secondary branches will show different lengths. Similarly, random elements will be added to the positioning points of the tertiary branches, rather than equal division of lengths, as shown in Figure 4 shown.

[0098] As the second part, leaf modeling is carried out:

[0099] To realistically represent the locust tree leaves and leaf clusters, 3ds Max modeling software is used to carry out single-leaf surface modeling, and the coordinates of the skeleton points and the mesh index between the points are obtained, as shown in Figure 5 shown. In the L-system simulation program, the point coordinates and indexes are read, and the single-leaf model is rendered and drawn, as shown in Figure 6 shown.

[0100] From the outlines of a batch of locust tree leaves, the outline curves of the sample leaves are extracted. During the modeling process, the sample is placed at the origin of coordinates. The x-axis and y-axis respectively simulate the lateral and longitudinal curvatures of the leaves. The sine function is used in the x direction of the lateral curvature and multiplied by a coefficient, which is a very small constant, aiming to reduce the amplitude of the entire curve shape (as shown in Figure 7 ). The cosine function is used in the y direction of the longitudinal curvature and multiplied by a coefficient, which is also a very small constant, with the same purpose of reducing the amplitude of the entire curve shape (as shown in Figure 8 ). The shape of the locust tree leaves in the natural state is restored as realistically as possible (as shown in Figure 9 ).

[0101] Due to the extremely large number of leaves on locust trees, to reduce the file size. After obtaining the Nurbs surface of the leaf morphology, it is converted into a mesh. In the process of converting the function surface to a mesh, the algorithm of converting by Surface sampling is adopted, and its core process and calculation formula are as follows:

[0102] 1) Map the points on the Nurbs surface to three-dimensional space and represent them as: z = f(x, y);

[0103] 2) Conduct mesh division and set the sampling area range [x min , x max and [y min , y max ;

[0104] 3) Determine the mesh resolution, including the number of sampling points in the x direction and y direction, which are N x and N y respectively, and calculate the sampling step in each direction as:

[0105]

[0106] 4) Generate all sampling points in the two-dimensional mesh by sampling the x value and y value, and represent them as:

[0107] x i = x min + i·Δx where i = 0, 1, ……, N x - 1;

[0108] y i = y min + j·Δy where j = 0, 1, ……, N y - 1;

[0109] 5) Generate the corresponding z value for the i-th sampling point (x i , y j ): z ij = f(x i , y j ), and combine the x value and y value to obtain the three-dimensional coordinates (x i , y j , z ij ) of each sampling point;

[0110] 6) Connect each sampling point to form a network. In the two-dimensional mesh, the sampling points form a mesh surface by forming quadrilaterals, and divide the small units of each mesh surface into two triangles:

[0111] The first triangle: (x i , yj , z ij ), (x i+1 , y j , z i+1,j ), (x i , y i+1 , z ij+1 );

[0112] The second triangle: (x i+1 , y j , z i+1,j ), (x i+1 , y i+1 , z i+1,j+1 ), (x i , y j+1 , z ij+1 );

[0113] 7) Since photosynthesis is considered when modeling the blade, the normal vector of each grid surface is calculated, and the formula is as follows:

[0114] n = P1 × P2;

[0115] In the formula, P1 and P2 are edge vectors derived from adjacent vertices. Using the vertex (x i , y j , z ij ) and adjacent vertices to calculate the normal vector, we get:

[0116]

[0117] Set corresponding rules to align the seam points, simplify the plane and refine the grid, and set the minimum and maximum numbers of the starting quadrilateral grid. According to the preset aspect ratio, set corresponding rules for the lengths of the largest and smallest edges to complete the grid conversion.

[0118] The grid after the first conversion is not the final required form of this embodiment, and grid reconstruction needs to be performed again.

[0119] To simplify the leaf morphological features as much as possible. During grid reconstruction, this embodiment selects quadrilateral grid reconstruction instead of triangular grid reconstruction (such as Figure 10 ). In the grid conversion algorithm, this embodiment sets the number of quadrilaterals and makes it adaptive. Maintain the original hard edges and set the radial structure line as the y-axis, so that the blade can not only maintain the simplified form and the curve radian in the x-axis and y-axis directions, but also maintain the simplest form.

[0120] When the model is displayed, this embodiment can display the grid in the simplest form of the leaf. When generating the drawing, this embodiment can use the grid subdivision algorithm to obtain the optimized natural-shaped leaves. On this basis, first observe the state of the natural locust tree branches. The leaves on both sides of the tip will appear in pairs.

[0121] When the leaf shape of the locust tree is placed on the branch. Simulate the influence of gravity on the topmost branchlet and make it bend. In addition, the length of the tip needs to be determined according to the size of the leaf. There will be no symmetric leaves growing within a certain range of the tip. Therefore, this embodiment intercepts a sub-curve on the original curve to obtain the bionic length value (such as Figure 11 ).

[0122] The growth rules of the locust tree leaves are different at the top and on both sides respectively. The leaves distributed on both sides are in symmetric positions, and the distance between them is approximately the width of the leaf, and a random value is added on this basis. As Figure 12 shown, where the green dots are the leaf positioning points after adding random value interference.

[0123] Finding the positioning points cannot fully satisfy the simulation of the natural state. Due to the existence of the random value parameter, the angle of the growth direction will be disrupted. In this state, it does not conform to the maximum photosynthetic utilization of plants. Similar to the sunflower, the sunflower will face its largest side towards the sun to obtain a longer sunshine duration and stronger sunshine intensity. As Figure 13 shown.

[0124] It is necessary to keep all the leaves in their original orientation and add the minimum value in the unified Z-axis direction. Find the plane with the smallest angle between the branch and the z-axis to maximize photosynthesis, that is, the sun-facing area can reach the maximum, which is most in line with the simulation of plant growth. Therefore, after finding the minimum plane. Calculate the normal vector, and then arrange all the leaves in sequence. The corresponding grid diagram is as Figure 14 shown.

[0125] Under normal conditions, the symmetric number of locust tree leaves: 2 - 12 pairs. Therefore, this embodiment will give a range when setting the length of the topmost branchlet, so that the coordinate points of the leaves are between 2 - 12 pairs. And a certain random value is given to interfere with the coordinate points of the leaves.

[0126] By comparing with the actual locust tree leaves, this embodiment finds that each locust tree leaf is of different sizes. In the natural growth state, the leaves in the middle part are larger, while the leaves on both sides are smaller. Therefore, when simulating the natural-shaped leaves, this embodiment adopts the method of gradually changing and randomly scaling the symmetric point positions to make each leaf different. As Figure 15 shown, where the green shape is the final shape of the leaf, and the red is the shape before optimization.

[0127] As the third part, root system modeling is carried out, such as Figure 20 shown: The shortest path algorithm and the slime mold algorithm can simulate the growth of the root system (the fluffiness of the variable soil quality can cause the root system to change from vertical growth to horizontal expansion).

[0128] The shortest path algorithm is used to find the shortest path between two nodes, which can be different soil conditions or obstacles encountered during the growth of the root system. The root system needs to find water and nutrients during growth, that is, the path of the root system from a certain point to the optimal growth point can be simulated by the shortest path algorithm; the slime mold algorithm mimics the growth behavior of slime molds in nature. This algorithm is mainly based on the network structure formed by slime molds when looking for food, simulates the adaptive behavior of organisms in nature, and explores and utilizes resources by establishing a dynamic network. Combining these two algorithms makes the model more realistic and efficient.

[0129] First, use the A* algorithm to find the shortest path from the root system to the target water source or nutrient source. The path can be represented by node records, and a shortest path is obtained; use the slime mold algorithm to grow dynamically in the grid, optimize for the distribution of water sources and nutrients, so as to adjust the root system structure; perform feedback regulation according to the growth of the root system affecting the environmental attributes; through multiple iterations of the A* and slime mold algorithms, continuously improve the adaptability of the root system and the acquisition efficiency of resources.

[0130] Steps of the A* algorithm: ① Create an open list and a closed list. The open list is used to store nodes to be evaluated, and the closed list is used to store evaluated nodes; ② Add the starting node to the open list; ③ Select the node with the lowest f(n)=g(n)+h(n) value from the open list, judge whether the node is the target node, if so, construct the path and exit, otherwise, move the current node to the closed list and check its neighboring nodes, and update the g and f values; ④ Finally, backtrack from the target node to the starting node to construct the path.

[0131] The A* algorithm performs path search through a heuristic method, and its formula is as follows:

[0132] f(n)=g(n)+h(n)

[0133]

[0134] In the formula, f(n): The total estimated cost of node n, used to evaluate whether this node should be expanded; g(n): The actual cost (path length) from the starting node to node n, usually obtained by accumulating the weights of the edges along the way; h(n): The heuristic estimated cost from node n to the target node; (x n ,y n ,z n ,) is the coordinate of node n, (x t, y t , z t ) are the coordinates of the target node.

[0135] Steps of the slime mold algorithm:

[0136] ① Create N slime mold individuals and randomly assign their initial positions in the grid

[0137] P i (0) = (x i (0), y i (0)) for i = 1, 2, ……, N

[0138] ② Calculate the fitness of each slime mold to judge the growth situation:

[0139] F i (t) = f(P i (t))

[0140] Wherein, f(P i (t)) = W[x i (t)][y i (t)] + λ·N[x i (t)][y i (t)].

[0141] ③ Update the positions of the slime mold individuals according to the current optimal position:

[0142] P i (t + 1) = P i (t) + α·dir(P best (t) - P i (t)) + β·rand(R)

[0143] ④ Repeat the position evaluation and update steps until the specified stop condition is reached, specifically:

[0144] if |F i (t) - F i (t - 1)| < ε or t >= t max then stop.

[0145] Finally, construct the locust tree model: Combine the branches and leaves to form the locust tree model of a single tree. The simulation result of a healthy growing locust tree is as Figure 16 shown. By setting different parameters, different forms of the tree can be obtained.

[0146] When simulating the dead branch rate, this embodiment uses three different filtering methods.

[0147] The first type of filtering is to filter at the height. The leaves with values ​​above the height are removed, and the leaves with values ​​below the height are retained (such as Figure 17 shown).

[0148] In the second filtering, the leaves that meet the height conditions are screened based on the leaves left by the first filtering. Randomly delete them in proportion and remove 30%. Taking the dead branch rate of 30% as an example, this embodiment removes all leaves above 7 meters. Its model is as follows Figure 18 shown.

[0149] The third filtering method is accomplished by spindle modeling. In natural form, when the dead branch rate is 30%, the whole tree is in the form of a spindle. This embodiment requires modeling the spindle of the screening rule.

[0150] Before generating leaves, this embodiment finds the minimum plane facing the z-axis of the sun. This plane can extract the normal vector, the starting point of the normal vector and the center point of the leaf in this embodiment.

[0151] The generated spindle is used as the outer frame boundary, and a computer is used to simulate whether all the center points are inside the spindle. This embodiment will obtain three answers, one is inside, one is outside, and one is on the surface of the spindle.

[0152] Because the growth rule of leaves is that the closer to the trunk, the less photosynthesis can be received, so there will be relatively fewer leaves near the trunk. In this embodiment, only the center points of leaves outside the spindle are removed, and a "non-strict" screening is performed.

[0153] The leaves that are retained at this point are the leaves that meet the set leaf withering rate under normal simulation conditions. The redundant branches are filtered out, and the final shape is as follows Figure 19 shown.

[0154] In order to speed up the operation, this embodiment simplifies the grid before calculating the volume and leaf area. By calculating the volume of the closed grid, accumulating and summing, and retaining two decimal places, the total volume of the trunk can be calculated to be 0.61 cubic meters. In the same way, in the leaves of the simplified grid, accumulating and summing, retaining two decimal places, the total area of ​​the leaves can be obtained to be 21.6 square meters. Figure 21 shown.

[0155] Therefore, this embodiment has the following beneficial effects:

[0156] (1) The object of study of this invention is the locust tree. The reason is that the locust tree, due to its drought tolerance, strong adaptability, and nitrogen fixation ability, helps to improve soil quality and enhance soil fertility, and can provide support for multiple fields such as ecological restoration, forestry management, and urban greening, promoting the coordinated development of society, economy, and environment. Therefore, constructing a locust tree model has extremely high value in actual production. Based on this, taking the L-system as the basic modeling method and adding various physical algorithms to construct the tree model, different morphological tree models can be constructed by adjusting parameters, so as to simulate the special growth mode morphology of tree species, and tree models with any degree of degradation can be constructed.

[0157] (2) When modeling the secondary branches of this invention, the tree species morphology such as the bending of branches caused by gravity and the different lengths of branches caused by the influence of photosynthesis are fully considered.

[0158] (3) When modeling the leaves of this invention, the following are considered: ① Constructing the leaf surface through a bending function to restore the leaf morphology of the locust tree in the natural state as much as possible. ② Determining the tip length according to the leaf size, the tip of the branch is bent due to the influence of gravity, the leaves at both ends of the branch are in symmetric positions, and a random value is added to make the growth spacing of leaf pairs different. ③ According to the maximum photosynthetic utilization of plants, while maintaining the original orientation and adding a unified minimum value in the Z direction, the minimum angle between the branch and the Z-axis is calculated to maximize the sun-facing area of the leaves, which is most in line with the plant growth mode.

[0159] (4) When modeling the root system of the locust tree of this invention, the shortest path algorithm and the slime mold algorithm are used to simulate the growth of the root system. The shortest path algorithm mainly generates the main path of the plant root system, while the slime mold algorithm imitates the adaptive behavior of organisms in nature to optimize the fine structure of the root system, making it more biologically meaningful and natural and realistic.

[0160] (5) After constructing the locust tree model of this invention, in order to enhance the authenticity of the model, improve the efficiency, highlight the structural characteristics of the tree, and meet the requirements of different application scenarios, dead tip filtering is carried out on the locust tree model. ① Height filtering: The leaves exceeding a certain height are removed, and the leaves below the height are retained. ② Random ratio filtering: On the basis of ①, according to the different dead tip rates of the locust tree in different seasons or different growth periods, the leaves can be deleted according to different ratios. ③ Spindle shape filtering: Through observation, in the natural state, when the dead tip rate of the locust tree is 30%, the morphology of the whole tree presents the shape of a spindle. Modeling the spindle of the screening rule, and using the generated spindle as the outer frame boundary to remove the leaves whose center points are outside the spindle.

[0161] (6) The model constructed by the present invention is highly simulated, and can model the models of healthy locust trees at any height. By comparing the parameters of the constructed locust tree model with the physical locust tree scanned by 3D point cloud technology and adding feedback regulation, the modeling accuracy is greatly improved. Moreover, it conforms to the growth law of specific tree species. Especially for the modeling of leaves, it can be extended to tree species with opposite compound leaf structures.

[0162] The various embodiments in this specification are described in a progressive manner. Each embodiment focuses on the differences from other embodiments. For the same or similar parts among the various embodiments, reference can be made to each other.

[0163] Specific examples are used in this article to elaborate on the principle and implementation manner of the present invention. The description of the above embodiments is only used to help understand the core idea of the present invention. At the same time, for those of ordinary skill in the art, according to the idea of the present invention, there will be changes in the specific implementation manner and application scope. In summary, the content of this specification should not be construed as a limitation on the present invention.

Claims

1. A modeling method for the whole process of black locust degradation based on the L-system, characterized in that, Including: Obtain the basic data of Robinia pseudoacacia; the basic data is determined from field measured data and literature materials; Extract L-system rules based on the basic data; the L-system rules include main trunk rules, branch pattern rules, and leaf simulation rules; Conduct modeling of the distribution of Robinia pseudoacacia trees according to the L-system rules to obtain main trunk and branch models, leaf models, and root models respectively; Combine a single Robinia pseudoacacia tree model based on the main trunk and branch models, the leaf model, and the root model, and simulate the dead tip rate of Robinia pseudoacacia trees to obtain a change model of the entire process of Robinia pseudoacacia degradation; The modeling process of the leaf model includes: Use 3ds Max modeling software to conduct single-leaf surface modeling, and determine the coordinates of the skeleton points and the network index between the points; Conduct trigonometric function modeling based on the coordinates of the skeleton points and the network index between the points to obtain a Nurbs surface of the leaf shape; Convert the Nurbs surface of the leaf shape into a mesh, and conduct quadrilateral mesh reconstruction, and adjust the growth shape of the leaves within the mesh to obtain the leaf model; The process of converting the Nurbs surface of the leaf shape into a mesh specifically includes: 1) Map the points on the Nurbs surface to three-dimensional space and represent them as: z = f(x, y); 2) Perform mesh generation and set the sampling area range [x min , x max and [y min , y max ; 3) Determine the grid resolution, including the number of sampling points in the x-direction and y-direction, which are N x and N y , and calculate the sampling step size in each direction as: 4) Generate all sampling points in the two-dimensional grid by sampling the x value and the y value, and represent them as: x i = x min + i·Δx where i = 0, 1, ……, N x -1; y i = y min + j·Δy where j = 0, 1, ……, N y - 1; 5) For the i-th sampling point (x i , y j ), generate the corresponding z value: z ij = f(x i , y j ), and combine the x value and the y value to obtain the three-dimensional coordinates (x i , y j , z ij ) of each sampling point; 6) Connect each sampling point to form a network. In the two-dimensional grid, the sampling points form a grid surface by forming quadrilaterals, and divide the small units of each grid surface into two triangles: The first triangle: (x i , y j , z ij ), (x i+1 , y j , z i+1,j ), (x i , y i+1 , z ij+1 ); Second triangle: (x i+1 , y j , z i+1,j ), (x i+1 , y i+1 , z i+1,j+1 ), (x i , y j+1 , z ij+1 ); 7) Since photosynthesis is considered when modeling the leaves, calculate the normal vector of each grid surface, and the formula is as follows: n = P1 × P2; Wherein, P1 and P2 are edge vectors derived from adjacent vertices, and the normal vector is calculated using the vertices (x i , y j , z ij ) and adjacent vertices, and we get: Set corresponding rules to align the seam points, and set the minimum and maximum numbers of the starting quadrilateral meshes, and set corresponding rules for the maximum and minimum edge lengths according to the preset aspect ratio to complete the mesh conversion.

2. The modeling method for the whole process of black locust degradation based on the L-system according to claim 1, characterized in that, The extraction formula of the L-system rules includes: The definition of the main trunk rule is the part below the branch height, which is expressed by the formula @!(w)F(l), where @ indicates that the following character is an axiom character,!(w) indicates that the radius of the main trunk drawn is w, and F(l) indicates that a frustum with a length of l is drawn as the branch internode; The branch pattern rule is expressed by the formula p1: A(l, w) →!(w)F(l)[&(a0)B(l × r2, w × wr)][ / (a1)B(l × r2, w × wr)][&(a0)B(l × r2, w × wr)][&(a1)B(l × r2, w × wr)], indicating that the vertex of the main axis A generates an upward internode F and four side branch vertices B in each step of derivation; where, l × r2 represents the length of the branch, w × wr represents the thickness of the branch, and &(a0), / (a1), &(a1) are all adjustment parameters of the branch morphological structure; The leaf simulation rule is expressed by the formula L(a, b, c), where a represents the branch segment length, b represents the number of leaves on the segment branch, and c represents the size of the leaves.

3. The modeling method for the whole process of black locust degradation based on the L-system according to claim 1, wherein, The modeling process of the main trunk and branch models includes: Derivation is carried out using the production A(l, w). The vertex of the main axis A generates an internode F and four lateral branches B at each step of the derivation, completing one round of branch growth, and multiple rounds of simulation of the natural growth structure of the branches are carried out according to the iteration step, gravity, and random length to obtain the main trunk and branch model.

4. The method for modeling the whole process of black locust degradation based on the L-system according to claim 1, wherein, The modeling process of the root system model includes: Using the A* algorithm to determine the shortest path from the root system to the target water source or nutrient source, and performing feedback adjustment according to the environmental attributes affecting the growth of the root system, and using the slime mold algorithm for dynamic growth to obtain the root system model.

5. The modeling method for the whole process of black locust degradation based on the L-system according to claim 1, characterized in that The simulation process of the withered tip rate of black locust trees includes height filtering, random ratio filtering, and spindle shape filtering; among them, the height filtering is to remove the leaves exceeding the set height and retain the leaves below the set height; the random ratio filtering is to delete the leaves according to the set ratio; the spindle shape filtering is to model the spindle shape of the set screening rules, and use the generated spindle shape model as the outer frame boundary of the single black locust tree model, and remove the leaves with the center point outside the boundary.

6. A modeling system for the whole process of black locust degradation based on the L-system, which applies the method described in any one of claims 1-5, characterized in that, It includes: A data acquisition unit for obtaining the basic data of black locust trees; the basic data is determined from field measured data and literature materials; A rule extraction unit for extracting L-system rules based on the basic data; the L-system rules include main trunk rules, branch pattern rules, and leaf simulation rules; A black locust tree branch modeling unit for performing black locust tree branch modeling according to the L-system rules to obtain the main trunk and branch model, leaf model, and root system model respectively; A black locust tree withered tip rate simulation unit for combining the single black locust tree model according to the main trunk and branch model, the leaf model, and the root system model, and simulating the withered tip rate of black locust trees to obtain the change model of the whole process of black locust degradation.

7. An electronic device, characterized in that, It includes a memory and a processor. The memory is used to store a computer program, and the processor runs the computer program to enable the electronic device to execute the method for modeling the whole process of black locust degradation based on the L-system according to any one of claims 1-5.

8. A computer-readable storage medium, characterized in that, It stores a computer program, and when the computer program is executed by the processor, it implements the method for modeling the whole process of black locust degradation based on the L-system according to any one of claims 1-5.

Images