An ecological garden greening design optimization method

Abstract

The application discloses an ecological garden greening design optimization method and relates to the technical field of garden greening design. By constructing an initial visual field characteristic vector FΩ, comprehensive extraction and integration of visual experience elements are realized. Further, by constructing an optimal vegetation height adjustment vector Hopt and combining a visual field obstacle matrix Um generated by fitting, not only can the key vegetation unit causing actual obstruction to the visual line of the visual point be accurately identified, but also the influence can be quantified as a viewing obstacle index U(k), and an optimized vegetation optimization density vector Pp is obtained accordingly, and finally a comprehensive visual field optimization index Q is calculated. The above processing process changes the traditional garden greening design relying on experience-based judgment into a systematic optimization process based on data driving and model reasoning, has obvious advantages in computability, structural expressiveness and visual smoothness, and significantly improves the scientificity and decision-making efficiency of the garden greening design scheme.

Description

Technical Field

[0001] This invention relates to the field of landscape design technology, specifically to an ecological landscape design optimization method. Background Technology

[0002] Ecological environment construction is a crucial component of modern urban development. Landscape planning, as a core area of ​​ecological construction, is dedicated to enhancing the ecological functions, aesthetic value, and sustainable development capabilities of urban green spaces. Within the sub-fields of landscape planning, ecological landscaping focuses on the rational allocation of natural elements such as vegetation, topography, and water bodies to improve the ecological stability and landscape diversity of green spaces. In this field, optimizing landscape views is increasingly becoming an important research direction, as it not only affects the viewing experience but also relates to the interaction between humans and nature. For example, in different settings such as parks, wetlands, urban greenways, and scenic areas, a well-designed landscape view can guide visitors' viewing routes, enhance the walking experience, and transform green spaces not only into part of the ecosystem but also into highly attractive cultural and recreational spaces.

[0003] Chinese Patent Application No. CN202311840936.0 discloses an ecological landscaping optimization method, specifically relating to the field of landscape design technology, comprising the following steps: Step 1, collecting target landscaping information and identifying green areas; Step 2, collecting regional information on the green areas from Step 1 and summarizing the collected information into a green area information set; Step 3, collecting green plant information and summarizing it into a green plant information set; Step 4, performing analysis operations based on the green area information set and the green plant information set to obtain suitable plant planting values ​​ZZi, and sorting them according to the suitable plant planting values ​​ZZi to obtain a suitable plant species ranking list. This invention can improve the ecological benefits of green areas. By collecting target landscaping information, identifying green areas, and collecting regional and green plant information, it is possible to better understand the ecological status of green areas and provide a scientific and reasonable basis for the selection and planting of green plants.

[0004] This reveals a significant deficiency in current ecological landscaping design regarding landscape visibility optimization. Traditional landscaping planning primarily focuses on vegetation adaptability, ecological restoration, and biodiversity, but pays less attention to visual transparency and landscape layering. In many urban greening and park landscape designs, trees, shrubs, and ground cover plants are often planted using uniform dense planting or random distribution patterns, lacking scientific adjustments based on human visual experience. This approach easily leads to visual obstruction problems, where some areas are blocked by overly dense trees or shrubs, limiting the viewing experience and making it difficult for visitors to appreciate the complete natural scenery. Furthermore, in some urban green spaces, parks, or scenic areas, the lack of layering in plant height configurations results in insufficient landscape depth, depriving people of the rich visual experience of varying distances. For areas requiring open views, such as lakes, mountains, and central park plazas, existing landscaping methods may obstruct some key attractions, weakening the aesthetic value of the garden space and limiting interaction between people and the environment. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides an ecological landscaping design optimization method, which solves the problems mentioned in the background section.

[0006] To achieve the above objectives, the present invention provides the following technical solution: an ecological landscape greening design optimization method, comprising the following steps:

[0007] S1. Collect three-dimensional topographic data and existing vegetation distribution data of the garden greening, form a topographic feature vector T and a vegetation feature matrix P, collect the observation point k of the garden greening, form a viewpoint dataset V, calculate the initial line-of-sight unobstructedness index Ω0, integrate the topographic feature vector T, vegetation feature matrix P, viewpoint dataset V and initial line-of-sight unobstructedness index Ω0, and obtain the initial field-of-sight feature vector FΩ.

[0008] S2. Calculate the optimal visible height of each observation point k in the viewpoint dataset V based on the initial field of view feature vector FΩ, and form the optimal vegetation height adjustment vector Hopt.

[0009] S3. Based on the obtained optimal vegetation height adjustment vector Hopt, it is then fitted with the terrain feature vector T and the vegetation feature matrix P to calculate the viewing obstacle index U(k) of each observation point k. After integration processing, the viewing obstacle matrix Um of the viewing point is obtained.

[0010] S4. By combining the obtained visual obstruction matrix Um and the optimal vegetation height adjustment vector Hopt, the vegetation planting density of the garden is optimized to obtain the vegetation optimization density vector Pp.

[0011] S5. Calculate the comprehensive field of vision optimization index Q by combining the vegetation optimization density vector Pp, the field of vision obstruction matrix Um, and the optimal vegetation height adjustment vector Hopt, and compare it with the preset optimization threshold Qth to determine whether to proceed to iterative optimization and decision execution.

[0012] Preferably, S1 includes S11, S12, S13 and S14;

[0013] S11. Real-time collection of three-dimensional topographic data and existing vegetation distribution data of the garden greening area through lidar mapping, GIS system and remote sensing imagery, forming a topographic feature vector T and a vegetation feature matrix P;

[0014] Wherein, the terrain feature vector T includes T={t1,t2,……,tn}, ti represents the feature vector of terrain point i, and n represents the total number of terrain feature points;

[0015] The feature vector ti of topographic point i specifically includes the elevation h(ti), slope s(ti), and surface reflectance λ(ti) of topographic point i;

[0016] The vegetation feature matrix P includes P = {hp(j), dp(j), λp(j)|j∈Mp}, where Mp represents the total amount of vegetation, hp(j) represents the height of vegetation j, dp(j) represents the canopy diameter of vegetation j, and λp(j) represents the leaf density of vegetation j.

[0017] Preferably, in step S12, observation points k of the garden greening are collected. Observation points k include roadsides, overpasses, pedestrian paths, viewing platforms, lakesides, and rest areas. The layer height of observation points k is extracted, including ground view, elevated view and distant landscape point. Then, the viewing angle range of observation points k between 30° and 120° is extracted to form the viewpoint dataset V.

[0018] Among them, the observation point collection includes using laser rangefinders, GIS systems and panoramic cameras to collect information on the hierarchical height, coordinates, viewing range and line of sight of the observation points;

[0019] The viewpoint dataset V includes V = {((xu(k), yu(k), hu(k), Ou(k), du(k))|k∈Mu}, where (xu(k), yu(k) represent the geographic coordinates of viewpoint k, hu(k) represent the height of viewpoint k, Ou(k) represent the viewing direction of viewpoint k, du(k) represent the maximum visible distance of viewpoint k; and Mu represents the total number of observation points.

[0020] Preferably, in step S13, the maximum visible area Atotal(k) and the initial occlusion area Ablock(k) of the observation point k are calculated based on the acquired viewpoint dataset V. By calculating the maximum visible area Atotal(k) and the initial occlusion area Ablock(k), the initial visual unobstructedness index Ω0 is obtained, which reflects the initial visual unobstructedness of the entire landscape greening.

[0021] The maximum visible area Attotal(k) is obtained using the following formula:

[0022] Atotal(k) = π*du(k) 2 ;

[0023] In the formula, π represents the constant of pi, with a value of 3.14;

[0024] The initial occlusion area Ablock(k) is obtained using the following formula:

[0025] Ablock(k) = ∑ j∈Mp f(hp(j),dp(j),λp(j),Ou(k));

[0026] In the formula, f represents the comprehensive function, which is specifically used to estimate the shading contribution of vegetation j to observation point k;

[0027] The initial visual unobstructedness index Ω0 is obtained using the following formula:

[0028]

[0029] In the formula, w(k) represents the weight coefficient of observation point k, which is preset according to the flow of people and functional level at observation point k;

[0030] S14. Based on the acquired terrain feature vector T, vegetation feature matrix P, viewpoint dataset V, and initial line-of-sight unobstructedness index Ω0, integrate them to obtain dataset V and initial line-of-sight unobstructedness index Ω0, and obtain the initial line-of-sight unobstructedness index Ω0.

[0031] Preferably, S2 includes S21;

[0032] S21. Based on the initial field of view feature vector FΩ, extract the viewpoint dataset V and vegetation feature matrix P. Based on the spatial relationship between the viewpoint dataset V and vegetation feature matrix P, estimate the optimal visible height required for unobstructed observation point k by weighted averaging the height, leaf density and spatial direction angle of vegetation within a specific angle range in front of observation point k, and construct the optimal vegetation height adjustment vector Hopt.

[0033] The optimal vegetation height adjustment vector Hopt is specifically Hopt={(Hopt(1), Hopt(1), …, Hopt(k))|k∈Mu};

[0034] The optimal vegetation height adjustment vector Hopt is obtained using the following formula:

[0035]

[0036] In the formula, Hopt(k) represents the optimal vegetation height adjustment vector at observation point k, P(k) represents the subset of vegetation feature matrices in the vegetation feature matrix P that have an occlusion relationship with observation point k, P(j)∈P(k) means that the vegetation feature matrix P of vegetation j belongs to the subset of the vegetation feature matrix P at observation point k, cos represents the cosine function, and φ(k,j) represents the directional angle between vegetation j and observation point k. When cos(φ(k,j))=0°, it means that vegetation j is directly in front of observation point k. When it deviates from the main direction, the directional angle φ(k,j) between vegetation j and observation point k changes adaptively, and the influence of vegetation j is adjusted synchronously and adaptively.

[0037] Preferably, S3 includes S31 and S32;

[0038] S31. Based on the obtained optimal vegetation height adjustment vector Hopt, the optimal vegetation height adjustment vector Hopt(k) at observation point k is extracted. Based on the terrain feature vector T and vegetation feature matrix P of the observation point k region, the viewing obstacle index U(k) at observation point k is calculated.

[0039] The viewing obstruction index U(k) at observation point k is obtained using the following formula:

[0040]

[0041] In the formula, U(k,j) represents the viewing obstacle index of vegetation j at observation point k, κ(k,j) represents the terrain shading factor, specifically representing the additional influence of terrain slope and elevation difference on vegetation j at observation point k, u1, u2 and u3 represent weighting factors, specifically u1 is used to balance the weighting factor of height exceeding the limit, u2 is used to balance the weighting factor of vegetation j density, u3 is used to balance the weighting factor of terrain shading, and u1+u2+u3=1, the specific value is set by the user.

[0042] Preferably, in step S32, based on the viewing obstacle index U(k) at the obtained observation point k, all observation points are summarized, and the specific occlusion contribution relationship between observation point k and vegetation j is represented in matrix form. The viewing obstacle matrix Um of the viewing point is obtained. The size of the viewing obstacle matrix Um is Mu*Mp. The occlusion intensity relationship between all observation points and vegetation is recorded.

[0043] The visual obstruction matrix Um has the following specific matrix form:

[0044]

[0045] Preferably, S4 includes S41;

[0046] S41. Extract the occlusion contribution of vegetation j to observation point k from the obtained visual obstruction matrix Um, and then combine it with the optimal vegetation height adjustment vector Hopt to calculate the density adjustment required for vegetation j relative to the occlusion intensity, optimize the vegetation planting density of the garden, and obtain the vegetation optimization density vector Pp.

[0047] The vegetation optimization density vector Pp is specifically Pp={Pp(1), Pp(2),……,Pp(j)|j∈Mp};

[0048] The vegetation optimization density vector Pp is obtained using the following formula:

[0049]

[0050] In the formula, Pp(j) represents the optimized planting density of vegetation j, Orig(j) represents the original planting density of vegetation j, which is set through the initial planning of the landscape greening, U(k,j) represents the viewing obstacle index of vegetation j at observation point k, hp(j) represents the height of vegetation j, and Hopt(k) represents the optimal vegetation height adjustment vector at observation point k.

[0051] Preferably, S5 includes S51 and S52;

[0052] S51. Combine the vegetation optimization density vector Pp, the visual obstruction matrix Um and the optimal vegetation height adjustment vector Hopt to calculate the comprehensive visual optimization index Q, which reflects the visual unobstructedness after the improvement of the landscape greening.

[0053] The overall field of view optimization index Q is obtained through the following formula:

[0054]

[0055] In the formula, Mp represents the total amount of vegetation, Mu represents the total number of observation points, and U(k,j) represents the viewing obstacle index of vegetation j at observation point k. This represents the mean value of vegetation obstruction. represents the vegetation height deviation, (1-Pp(j)) represents the density reduction term, q1, q2 and q3 represent the weight coefficients of the average vegetation obstruction, vegetation height deviation and density reduction term respectively, and q1+q2+q3=1, the specific values ​​are set by the user.

[0056] Preferably, in step S52, the obtained comprehensive vision optimization index Q is compared with the preset optimization threshold Qth to obtain the comparison result, and the process of entering iterative optimization and decision execution is determined based on the comparison result.

[0057] The comparison results were obtained through the following comparison methods:

[0058] When the comprehensive field of view optimization index Q ≤ optimization threshold Qth, it means that the comparison result is passed and the iterative optimization stage is not entered. Instead, the decision execution stage is entered, which includes generating vegetation height adjustment suggestions based on the optimal vegetation height adjustment vector Hopt, generating vegetation density adjustment suggestions based on the vegetation optimization density vector Pp, and generating occlusion observation point intervention suggestions based on the field of view obstacle matrix Um.

[0059] When the comprehensive vision optimization index Q < optimization threshold Qth, it indicates that the comparison result is not passed, and the system will not enter the decision execution environment, but will enter the iterative optimization stage, including executing S2 to S4 for iterative adjustment.

[0060] This invention provides an optimized method for ecological landscaping design, which has the following beneficial effects:

[0061] (1) By constructing an initial visual feature vector FΩ, a comprehensive extraction and integration of factors affecting visual experience is achieved. Furthermore, by constructing an optimal vegetation height adjustment vector Hopt, combined with the fitted visual obstruction matrix Um of the viewpoint, not only can key vegetation units that actually obstruct the viewpoint's line of sight be accurately identified, but this influence can also be quantified into a view obstruction index U(k). Based on this, an optimized vegetation density vector Pp is obtained, and finally, a comprehensive visual optimization index Q is calculated. This process transforms landscape design from a traditional reliance on empirical judgment to a system optimization process based on data-driven and model-based reasoning, effectively solving the common problems in existing landscape design such as "lack of quantitative support for unobstructed visual experience," "imbalance between green density and landscape value," and "difficulty in dynamically judging the quality of green configuration." Compared to traditional methods relying on human experience, this invention has significant advantages in computability, structural expressiveness, and improved visual unobstructedness, significantly improving the scientific nature and decision-making efficiency of landscape design schemes.

[0062] (2) Through the collection and modeling of three-dimensional terrain and multi-dimensional vegetation features in the garden green area, a multi-dimensional structured data foundation was constructed, including terrain feature vector T, vegetation feature matrix P, and viewpoint dataset V. In particular, by comprehensively recording the viewing angle range, viewpoint height, and maximum visible distance of observation point k under different scene conditions, the viewpoint dataset V can truly reflect the actual visual contact paths of people in the park at multiple levels and in multiple directions. At the same time, by comparing the maximum visible area Atotal(k) with the initial occlusion area Ablock(k), and combining the spatial distribution characteristics of vegetation and directional occlusion relationship, an initial visual accessibility index Ω0 based on the observation point was established for the first time. This makes the accessibility index not only have an area geometric basis, but also have the ability to estimate the weight of vegetation physical attributes. Based on this, by obtaining the initial visual feature vector FΩ, a comprehensive modeling data foundation with spatial accuracy, angular relationship and ecological parameters is provided for subsequent visual optimization model calculation. This breaks through the core bottleneck of "data isolation, weak spatial correlation and inability to quantitatively evaluate initial permeability" in traditional greening layout, and significantly improves the data integrity and scientific starting point of analysis before landscape design optimization.

[0063] (3) By constructing a comprehensive field-of-view optimization index Q, a multi-factor collaborative evaluation mechanism is realized in the optimization process of landscape design. This mechanism can simultaneously integrate three key factors: spatial density regulation, shading intensity, and height control, as reflected by the vegetation optimization density vector Pp, the field-of-view obstruction matrix Um, and the optimal vegetation height adjustment vector Hopt. A data-driven stage-based judgment strategy is introduced into the landscape design optimization process. This not only avoids ineffective and repetitive subjective judgments but also allows for entry into the decision-making execution stage or automatic triggering of iterative optimization processes based on comparison results, forming a complete closed loop of optimization → evaluation → feedback → decision. This mechanism significantly enhances the intelligent response capability of landscape design methods, enabling them to maintain continuous optimization capabilities and controllable goal achievement in dynamically changing or complex multi-source environments. Attached Figure Description

[0064] Figure 1 This is a schematic diagram illustrating the steps of an ecological landscape greening design optimization method according to the present invention;

[0065] Figure 2 This is a schematic diagram showing the density change trend of some vegetation numbers under multiple rounds of optimization. Detailed Implementation

[0066] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0067] Example 1

[0068] This invention provides an optimization method for ecological landscaping design. Please refer to [link / reference]. Figure 1 This includes the following steps:

[0069] S1. Collect three-dimensional topographic data and existing vegetation distribution data of the garden greening, form a topographic feature vector T and a vegetation feature matrix P, collect the observation point k of the garden greening, form a viewpoint dataset V, calculate the initial line-of-sight unobstructedness index Ω0, integrate the topographic feature vector T, vegetation feature matrix P, viewpoint dataset V and initial line-of-sight unobstructedness index Ω0, and obtain the initial field-of-sight feature vector FΩ.

[0070] S2. Calculate the optimal visible height of each observation point k in the viewpoint dataset V based on the initial field of view feature vector FΩ, and form the optimal vegetation height adjustment vector Hopt.

[0071] S3. Based on the obtained optimal vegetation height adjustment vector Hopt, it is then fitted with the terrain feature vector T and the vegetation feature matrix P to calculate the viewing obstacle index U(k) of each observation point k. After integration processing, the viewing obstacle matrix Um of the viewing point is obtained.

[0072] S4. By combining the obtained visual obstruction matrix Um and the optimal vegetation height adjustment vector Hopt, the vegetation planting density of the garden is optimized to obtain the vegetation optimization density vector Pp.

[0073] S5. Calculate the comprehensive field of vision optimization index Q by combining the vegetation optimization density vector Pp, the field of vision obstruction matrix Um, and the optimal vegetation height adjustment vector Hopt, and compare it with the preset optimization threshold Qth to determine whether to proceed to iterative optimization and decision execution.

[0074] In this embodiment, by constructing a terrain feature vector T, a vegetation feature matrix P, a viewpoint dataset V, and an initial visual unobstructedness index ΩO, an initial visual field feature vector FΩ is formed, achieving comprehensive extraction and integration of elements affecting visual experience. Furthermore, by constructing an optimal vegetation height adjustment vector Hopt, combined with the fitted viewpoint visual obstruction matrix Um, not only can key vegetation units that actually obstruct the viewpoint's line of sight be accurately identified, but this influence can also be quantified into a viewpoint obstruction index U(k), and based on this, an optimized vegetation density vector Pp is obtained, ultimately calculating the comprehensive visual field optimization index Q. This process transforms landscape design from a traditional reliance on empirical judgment to a system optimization process based on data-driven and model-based reasoning, effectively solving the common problems in existing landscape design such as "lack of quantitative support for visual unobstructed experience," "imbalance between green density and landscape value," and "difficulty in dynamically judging the quality of greening configuration." In particular, this invention achieves automatic iterative optimization and adjustment decision output of the overall greening system through a comparison and control mechanism of the comprehensive vision optimization index Q and the optimization threshold Qth. This enables the landscape design to possess calculable, evaluable, and closed-loop adaptive control capabilities, enhancing visual transparency and aesthetics while also considering ecological functions and planting rationality. It is highly systematic and innovative.

[0075] Compared to traditional methods that rely on human experience, this invention has significant advantages in terms of computability, structural expressiveness, and visual fluency, thereby significantly improving the scientific nature and decision-making efficiency of landscape design schemes.

[0076] Example 2

[0077] This embodiment is an explanation based on Embodiment 1. Please refer to it. Figure 1 Specifically: S1 includes S11, S12, S13 and S14;

[0078] S11. Real-time collection of three-dimensional topographic data and existing vegetation distribution data of the garden greening area through lidar mapping, GIS system and remote sensing imagery, forming a topographic feature vector T and a vegetation feature matrix P;

[0079] Wherein, the terrain feature vector T includes T={t1,t2,……,tn}, ti represents the feature vector of terrain point i, and n represents the total number of terrain feature points;

[0080] The feature vector ti of topographic point i specifically includes the elevation h(ti), slope s(ti), and surface reflectance λ(ti) of topographic point i;

[0081] The vegetation feature matrix P includes P = {hp(j), dp(j), λp(j)|j∈Mp}, where Mp represents the total amount of vegetation, hp(j) represents the height of vegetation j, dp(j) represents the canopy diameter of vegetation j, and λp(j) represents the leaf density of vegetation j.

[0082] S12. Collect observation points k of the garden greening. Observation points k include roadsides, overpasses, pedestrian paths, viewing platforms, lakesides and rest areas. Extract the layer height of observation points k, including ground view, elevated view and distant landscape point. Then extract the viewing angle range of observation points k between 30° and 120° to form viewpoint dataset V.

[0083] The settings for ground view, elevated view, and distant viewpoint are usually as follows: ground view (1.5m), elevated view (3m-5m), and distant viewpoint (10m+);

[0084] Among them, the observation point collection includes using laser rangefinders, GIS systems and panoramic cameras to collect information on the hierarchical height, coordinates, viewing range and line of sight of the observation points;

[0085] The viewpoint dataset V includes V = {((xu(k), yu(k), hu(k), Ou(k), du(k))|k∈Mu}, where (xu(k), yu(k) represent the geographic coordinates of viewpoint k, hu(k) represent the height of viewpoint k, Ou(k) represent the viewing direction of viewpoint k, specifically representing the main visible range of viewpoint k, du(k) represent the maximum visible distance of viewpoint k, specifically representing the farthest green area that viewpoint k can observe; Mu represents the total number of observation points.

[0086] S13. Based on the acquired viewpoint dataset V, calculate the maximum visible area Atotal(k) and the initial occlusion area Ablock(k) of the observation point k. By calculating the maximum visible area Atotal(k) and the initial occlusion area Ablock(k), obtain the initial visual unobstructedness index Ω0, which reflects the overall initial visual unobstructedness of the garden greening.

[0087] The maximum visible area Attotal(k) is obtained using the following formula:

[0088] Atotal(k) = π*du(k) 2 ;

[0089] In the formula, π represents the constant of pi, with a value of 3.14. The significance of this formula is that, with the observation point k as the center and the maximum visible distance as the radius, an ideal unobstructed circular area is simulated, which is used to establish a comparison benchmark for "how much is obstructed" in the future.

[0090] The initial occlusion area Ablock(k) is obtained using the following formula:

[0091] Ablock(k) = ∑ j∈Mp f(hp(j),dp(j),λp(j),Ou(k));

[0092] In the formula, f represents the comprehensive function, which is used to estimate the occlusion contribution of vegetation j to observation point k. The comprehensive function f is usually obtained by modeling and fitting experimental data, and is taken as: f=α*hp(j)*dp(j)*λp(j)*δ(Op(j),Ou(k)), where Op(j) represents the angle of vegetation j, specifically the angle between the center of vegetation j and the direction of observation point k, and δ represents the view matching function, which determines whether vegetation j is within the view range of observation point k, and the output is 0 and 1. The significance of this formula is that by traversing all vegetation, combining their physical characteristics with the spatial relationship of observation point k, the actual occluded field of view area of ​​each observation point k is estimated for subsequent unobstructedness assessment.

[0093] The initial visual unobstructedness index Ω0 is obtained using the following formula:

[0094]

[0095] In the formula, w(k) represents the weight coefficient of observation point k, which is preset according to the flow of people and functional level at observation point k;

[0096] S14. Based on the acquired terrain feature vector T, vegetation feature matrix P, viewpoint dataset V, and initial line-of-sight unobstructedness index Ω0, integrate them to obtain dataset V and initial line-of-sight unobstructedness index Ω0, and obtain the initial line-of-sight unobstructedness index Ω0.

[0097] In this embodiment, by collecting and modeling three-dimensional terrain and multi-dimensional vegetation features of the garden green area, a multi-dimensional structured data foundation was constructed, including terrain feature vector T, vegetation feature matrix P, and viewpoint dataset V. In particular, by comprehensively recording the viewing angle range, viewpoint height, and maximum visible distance of observation point k under different scene conditions (such as ground, elevated, and distant views), the viewpoint dataset V can realistically reflect the actual visual contact paths of people in the park at multiple levels and in multiple directions. At the same time, by comparing the maximum visible area Atotal(k) with the initial occlusion area Ablock(k), and combining the spatial distribution characteristics of vegetation and directional occlusion relationships, an initial visual accessibility index Ω0 based on the observation point was established for the first time. This makes the accessibility index not only have an area geometric basis but also the ability to estimate the weights of vegetation physical attributes. Based on this, by obtaining the initial visual feature vector FΩ, a comprehensive modeling data foundation with spatial accuracy, angular relationship and ecological parameters is provided for subsequent visual optimization model calculation. This breaks through the core bottleneck of "data isolation, weak spatial correlation and inability to quantitatively evaluate initial permeability" in traditional greening layout, and significantly improves the data integrity and scientific starting point of analysis before landscape design optimization.

[0098] Example 3

[0099] This embodiment is an explanation based on Embodiment 2. Please refer to it. Figure 1 Specifically: S2 includes S21;

[0100] S21. Based on the initial field of view feature vector FΩ, extract the viewpoint dataset V and vegetation feature matrix P. Based on the spatial relationship between the viewpoint dataset V and vegetation feature matrix P, estimate the optimal visible height required for unobstructed observation point k by weighted averaging the height, leaf density and spatial direction angle of vegetation within a specific angle range in front of observation point k, and construct the optimal vegetation height adjustment vector Hopt.

[0101] The optimal vegetation height adjustment vector Hopt is specifically Hopt={(Hopt(1), Hopt(1), …, Hopt(k))|k∈Mu};

[0102] The optimal vegetation height adjustment vector Hopt is obtained using the following formula:

[0103]

[0104] In the formula, Hopt(k) represents the optimal vegetation height adjustment vector at observation point k, P(k) represents the subset of vegetation feature matrices in the vegetation feature matrix P that have an occlusion relationship with observation point k, P(j)∈P(k) means that the vegetation feature matrix P of vegetation j belongs to the subset of the vegetation feature matrix P at observation point k, cos represents the cosine function, φ(k,j) represents the directional angle between vegetation j and observation point k. When cos(φ(k,j))=0°, it means that vegetation j is directly in front of observation point k. When it deviates from the main direction, the directional angle φ(k,j) between vegetation j and observation point k changes adaptively, and the influence of vegetation j is adjusted synchronously and adaptively.

[0105] Table 1 shows the subset of vegetation feature matrix P corresponding to observation point k: {P1, P2, P3}.

[0106]

[0107] The calculated value of Hopt(k) is 1 / 3*{5.0*0.8*0.9659+6.0*0.6*0.8192+4.5*0.7*0.5}≈2.80;

[0108] The ideal maximum vegetation height in the area corresponding to observation point k is approximately 2.80.

[0109] S3 includes S31 and S32;

[0110] S31. Based on the obtained optimal vegetation height adjustment vector Hopt, the optimal vegetation height adjustment vector Hopt(k) at observation point k is extracted. Based on the terrain feature vector T and vegetation feature matrix P of the observation point k region, the viewing obstacle index U(k) at observation point k is calculated.

[0111] The viewing obstruction index U(k) at observation point k is obtained using the following formula:

[0112]

[0113] In the formula, U(k,j) represents the viewing obstacle index of vegetation j at observation point k, κ(k,j) represents the terrain shading factor, specifically representing the additional influence of terrain slope and elevation difference on vegetation j at observation point k, u1, u2 and u3 represent weighting factors, specifically u1 is used to balance the weighting factor of height exceeding the limit, u2 is used to balance the weighting factor of vegetation j density, u3 is used to balance the weighting factor of terrain shading, and u1+u2+u3=1, the specific value is set by the user.

[0114] In this embodiment, by leveraging the spatial correlation between the viewpoint dataset V extracted from the initial field-of-view feature vector FΩ and the vegetation feature matrix P, the vegetation height, leaf density, and directional angle within a specific angular range in front of each observation point k are weighted and fused to generate the optimal vegetation height adjustment vector Hopt. This vector dynamically reflects the ideal unobstructed viewing height requirements under different positions and observation angles. Especially in areas with overlapping multi-angle fields of view, it effectively captures subtle changes in the influence of occlusion from different directions, improving the angular resolution capability for judging regional visibility. Based on this, by jointly fitting Hopt(k) with the terrain feature vector T and vegetation feature matrix P within the area of ​​observation point k, a viewing obstacle index U(k) is constructed. A comprehensive evaluation factor, including height deviation, vegetation leaf density, and terrain slope influence, is introduced. This makes the occlusion judgment not only limited to whether it is visible within the visual range but also refined to "the specific occlusion intensity under the current height expectation and terrain interference conditions," significantly enhancing the sensitivity to field-of-view occlusion in micro-topographical environments and complex planting structures. This process can provide a robust basis for spatial occlusion quantification for subsequent density adjustment and optimization, and solves the problem that traditional methods are difficult to accurately perceive the degree of local visual interference under irregular terrain and dynamic field of view. It has stronger regional adaptability and structural recognition capabilities.

[0115] Example 4

[0116] This embodiment is an explanation based on Embodiment 3. Please refer to it. Figure 1 Specifically: S32. Based on the viewing obstacle index U(k) at the obtained observation point k, summarize all observation points and represent the specific shading contribution relationship between observation point k and vegetation j in matrix form. Obtain the viewing obstacle matrix Um of the viewing point. The size of the viewing obstacle matrix Um is Mu*Mp. Record the shading intensity relationship between all observation points and vegetation.

[0117] The visual obstruction matrix Um has the following specific matrix form:

[0118]

[0119] Each matrix element is specifically... In the formula, (.) + This means that if the value is less than 0, it will be 0 to prevent negative numbers from having an effect.

[0120] S4 includes S41;

[0121] S41. Extract the occlusion contribution of vegetation j to observation point k from the obtained visual obstruction matrix Um, and then combine it with the optimal vegetation height adjustment vector Hopt to calculate the density adjustment required for vegetation j relative to the occlusion intensity, optimize the vegetation planting density of the garden, and obtain the vegetation optimization density vector Pp.

[0122] The vegetation optimization density vector Pp is specifically Pp={Pp(1), Pp(2),……,Pp(j)|j∈Mp};

[0123] The vegetation optimization density vector Pp is obtained using the following formula:

[0124]

[0125] In the formula, Pp(j) represents the optimized planting density of vegetation j, Orig(j) represents the original planting density of vegetation j, which is set through the initial planning of the landscape greening, U(k,j) represents the viewing obstacle index of vegetation j at observation point k, hp(j) represents the height of vegetation j, and Hopt(k) represents the optimal vegetation height adjustment vector at observation point k.

[0126] In this embodiment, based on the constructed visual obstruction matrix Um, the pairwise occlusion contribution intensity U(k,j) between observation point k and vegetation j can be comprehensively recorded in a matrix structure. This not only provides fine-grained visual impact data but also breaks through the limitation of traditional landscape design, which can only evaluate global average occlusion. For the first time, it realizes explicit quantification of occlusion relationships between "multiple viewpoints → multiple vegetation", possessing stronger traceability and local controllability of occlusion. On this basis, by combining the height deviation relationship between the optimal vegetation height adjustment vector Hopt(k) and the actual vegetation height hp(j) and superimposing the occlusion contribution coefficient in the visual obstruction matrix Um, the vegetation optimization density vector Pp is dynamically adjusted and generated based on the original planting density Orig(j). This makes density optimization no longer limited to the average processing of the entire area but responsively adjusted for the specific occlusion performance and expected height deviation of each vegetation unit, significantly enhancing the matching accuracy between planting density and visual quality during the greening layout process. This mechanism provides a comprehensive optimization method for actual landscape configuration, which has the ability to identify interference, respond to targets, and coordinate density control. While maintaining the ecological effect of green space, it can achieve differentiated control targets for local visibility.

[0127] Example 5

[0128] This embodiment is an explanation based on Embodiment 4. Please refer to it. Figure 1 Specifically: S5 includes S51 and S52;

[0129] S51. Combine the vegetation optimization density vector Pp, the visual obstruction matrix Um and the optimal vegetation height adjustment vector Hopt to calculate the comprehensive visual optimization index Q, which reflects the visual unobstructedness after the improvement of the landscape greening.

[0130] The overall field of view optimization index Q is obtained through the following formula:

[0131]

[0132] In the formula, Mp represents the total amount of vegetation, Mu represents the total number of observation points, and U(k,j) represents the viewing obstacle index of vegetation j at observation point k. This represents the average shading value of vegetation, specifically indicating the degree of negative impact of the vegetation on the overall view of the park. The value represents the vegetation height deviation, specifically whether the overall height of the vegetation is still higher than the recommended value from different viewpoints. The larger the value, the more "too tall" the tree is. (1-Pp(j)) represents the density reduction level. Specifically, if the density has been adjusted to the minimum (e.g., 0), this item is 1, indicating that the adjustment is in place. If it is still the original density (e.g., 1), this item is 0, indicating that it has not been adjusted. q1, q2, and q3 represent the weighting coefficients of the average vegetation obstruction, vegetation height deviation, and density reduction level, respectively, and q1+q2+q3=1. The specific values ​​are set by the user.

[0133] S52. Compare the obtained comprehensive vision optimization index Q with the preset optimization threshold Qth, obtain the comparison result, and determine whether to proceed to iterative optimization and decision execution based on the comparison result;

[0134] The comparison results were obtained through the following comparison methods:

[0135] When the comprehensive field of view optimization index Q ≤ optimization threshold Qth, it means that the comparison result is passed and the iterative optimization stage is not entered. Instead, the decision execution stage is entered, which includes generating vegetation height adjustment suggestions based on the optimal vegetation height adjustment vector Hopt, generating vegetation density adjustment suggestions based on the vegetation optimization density vector Pp, and generating occlusion observation point intervention suggestions based on the field of view obstacle matrix Um.

[0136] When the comprehensive vision optimization index Q < optimization threshold Qth, it indicates that the comparison result is not passed, and the system will not enter the decision execution environment, but will enter the iterative optimization stage, including executing S2 to S4 for iterative adjustment.

[0137] In this embodiment, a multi-factor collaborative evaluation mechanism is realized in the optimization process of landscape design by constructing a comprehensive view optimization index Q. This mechanism can simultaneously integrate three key factors: spatial density regulation, shading intensity, and height control, reflected by the vegetation optimization density vector Pp, the view obstruction matrix Um, and the optimal vegetation height adjustment vector Hopt. The comprehensive view optimization index Q is composed of the average vegetation obstruction value, vegetation height deviation, and density reduction degree. It not only accurately reflects the overall improvement in the park's view but also allows for flexible setting of weighting coefficients q1, q2, and q3 to adapt to different design objectives regarding permeability, aesthetics, or ecological coverage, thus enhancing the method's parameter control adaptability. Furthermore, the automatic comparison mechanism between the comprehensive view optimization index Q and the optimization threshold Qth introduces a data-driven stage judgment strategy into the landscape design optimization process. This not only avoids ineffective and repetitive subjective judgments but also allows for entry into the decision-making execution stage or automatic triggering of iterative optimization processes based on the comparison results, forming a complete closed loop of optimization → evaluation → feedback → decision. This mechanism significantly enhances the intelligent response capability of landscape design methods, enabling them to maintain continuous optimization capabilities and controllability in achieving objectives in dynamically changing or complex multi-source environments.

[0138] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for optimizing ecological landscape greening design, characterized in that: Includes the following steps: S1. Collect three-dimensional topographic data and existing vegetation distribution data of the garden greening, form a topographic feature vector T and a vegetation feature matrix P, collect the observation points k of the garden greening, form a viewpoint dataset V, calculate the initial line-of-sight index Ω0, integrate the topographic feature vector T, vegetation feature matrix P, viewpoint dataset V and initial line-of-sight index Ω0, and obtain the initial field-of-sight feature vector FΩ. S1 includes S11 and S12; S11. Real-time collection of three-dimensional topographic data and existing vegetation distribution data of the garden greening area through lidar mapping, GIS system and remote sensing imagery, forming a topographic feature vector T and a vegetation feature matrix P; The terrain feature vector T includes T = {t1, t2, ..., ti, tn}, where ti represents the feature vector of terrain point i and n represents the total number of terrain feature points. The feature vector ti of topographic point i specifically includes the elevation h (ti), slope s (ti), and surface reflectance λ (ti) at topographic point i. The vegetation feature matrix P includes P={hp(j),dp(j),λp(j)|j∈Mp}, where Mp represents the total number of vegetation, hp(j) represents the height of vegetation j, dp(j) represents the canopy diameter of vegetation j, and λp(j) represents the leaf density of vegetation j. S12. Collect observation points k of the garden greening. Observation points k include roadsides, overpasses, pedestrian paths, viewing platforms, lakesides and rest areas. Extract the layer height of observation points k, including ground view, elevated view and distant landscape point. Then extract the viewing angle range of observation points k between 30° and 120° to form viewpoint dataset V. Among them, the observation point collection includes using laser rangefinders, GIS systems and panoramic cameras to collect information on the hierarchical height, coordinates, viewing range and line of sight of the observation points; The viewpoint dataset V includes V = {(xu(k), yu(k), hu(k), Ou(k), du(k)) | k∈Mu}, where (xu(k), yu(k)) represent the geographic coordinates of viewpoint k, hu(k) represent the height of viewpoint k, Ou(k) represent the viewing direction of viewpoint k, du(k) represent the maximum visible distance of viewpoint k, and Mu represents the total number of observation points; S2. Calculate the optimal visible height of each observation point k in the viewpoint dataset V based on the initial field of view feature vector FΩ, and form the optimal vegetation height adjustment vector Hopt. S3. Based on the obtained optimal vegetation height adjustment vector Hopt, it is then fitted with the terrain feature vector T and the vegetation feature matrix P to calculate the viewing obstacle index U(k) of each observation point k. After integration processing, the viewing obstacle matrix Um of the viewing point is obtained. S4. By combining the obtained visual obstruction matrix Um and the optimal vegetation height adjustment vector Hopt, the vegetation planting density of the garden is optimized to obtain the vegetation optimization density vector Pp. S5. Calculate the comprehensive field of vision optimization index Q by combining the vegetation optimization density vector Pp, the field of vision obstruction matrix Um, and the optimal vegetation height adjustment vector Hopt, and compare it with the preset optimization threshold Qth to determine whether to proceed to iterative optimization and decision execution.

2. The method for optimizing ecological landscaping design according to claim 1, characterized in that: S1 also includes S13 and S14; S13. Based on the acquired viewpoint dataset V, calculate the maximum visible area Atotal(k) and the initial occlusion area Ablock(k) of the observation point k. By calculating the maximum visible area Atotal(k) and the initial occlusion area Ablock(k), obtain the initial visual unobstructedness index Ω0, which reflects the overall initial visual unobstructedness of the garden greening. The maximum visible area Attotal(k) is obtained using the following formula: ； In the formula, π represents the constant of pi, with a value of 3.14; The initial occlusion area Ablock(k) is obtained using the following formula: ； In the formula, f represents the comprehensive function, which is specifically used to estimate the shading contribution of vegetation j to observation point k; The initial visual unobstructedness index ΩO is obtained using the following formula: ； In the formula, w(k) represents the weight coefficient of observation point k, which is preset according to the flow of people and functional level at observation point k; S14. Based on the acquired terrain feature vector T, vegetation feature matrix P, viewpoint dataset V, and initial line-of-sight unobstructedness index Ω0, the initial field-of-sight feature vector FΩ is obtained by integrating them.

3. The method for optimizing ecological landscaping design according to claim 2, characterized in that: S2 includes S21; S21. Based on the initial field of view feature vector FΩ, extract the viewpoint dataset V and vegetation feature matrix P. Based on the spatial relationship between the viewpoint dataset V and vegetation feature matrix P, estimate the optimal visible height required for unobstructed observation point k by weighted averaging the height, leaf density and spatial direction angle of vegetation within a specific angle range in front of observation point k, and construct the optimal vegetation height adjustment vector Hopt. The optimal vegetation height adjustment vector Hopt is specifically Hopt={(Hopt(1),Hopt(2),…,Hopt(k))|k∈Mu}; The optimal vegetation height adjustment vector Hopt is obtained using the following formula: ； In the formula, Hopt(k) represents the optimal vegetation height adjustment vector at observation point k, P(k) represents the subset of vegetation feature matrices in the vegetation feature matrix P that have an occlusion relationship with observation point k, p(j)∈P(k) means that the vegetation feature matrix P of vegetation j belongs to the subset of the vegetation feature matrix P at observation point k, and cos represents the cosine function. This represents the directional angle between vegetation j and observation point k, when When = 0, it indicates that vegetation j is directly in front of observation point k. When deviating from the main direction of realization, the angle between the directions of vegetation j and observation point k is... Adaptive change, synchronous adaptive adjustment of vegetation j.

4. The method for optimizing ecological landscaping design according to claim 3, characterized in that: S3 includes S31 and S32; S31. Based on the obtained optimal vegetation height adjustment vector Hopt, the optimal vegetation height adjustment vector Hopt(k) at observation point k is extracted. Based on the terrain feature vector T and vegetation feature matrix P of the observation point k region, the viewing obstacle index U(k) at observation point k is calculated. The viewing obstruction index U(k) at observation point k is obtained using the following formula: ； In the formula, U(k,j) represents the viewing obstruction index of vegetation j at observation point k. The topographic shading factor represents the additional influence of topographic slope and elevation difference on vegetation j at observation point k. u1, u2, and u3 represent weighting factors. Specifically, u1 is used to balance the weighting factor for excessive height, u2 is used to balance the weighting factor for vegetation j density, and u3 is used to balance the weighting factor for topographic shading. u1+u2+u3=1. The specific values ​​are set by the user.

5. The method for optimizing ecological landscaping design according to claim 4, characterized in that: S32. Based on the viewing obstacle index U(k) at the obtained observation point k, summarize all observation points and represent the specific shading contribution relationship between observation point k and vegetation j in matrix form. Obtain the viewing obstacle matrix Um of the viewing point. The size of the viewing obstacle matrix Um is Mu*Mp. Record the shading intensity relationship between all observation points and vegetation. The visual obstruction matrix Um has the following specific matrix form: 。 6. The method for optimizing ecological landscaping design according to claim 5, characterized in that: S4 includes S41; S41. Extract the occlusion contribution of vegetation j to observation point k from the obtained visual obstruction matrix Um, and then combine it with the optimal vegetation height adjustment vector Hopt to calculate the density adjustment required for vegetation j relative to the occlusion intensity, optimize the vegetation planting density of the garden, and obtain the vegetation optimization density vector Pp. The vegetation optimization density vector Pp is specifically Pp={Pp(1),Pp(2),……,Pp(j)|j∈Mp}; The vegetation optimization density vector Pp is obtained using the following formula: ； In the formula, Pp(j) represents the optimized planting density of vegetation j, Orig(j) represents the original planting density of vegetation j, which is set through the initial planning of the landscape greening, U(k,j) represents the viewing obstacle index of vegetation j at observation point k, hp(j) represents the height of vegetation j, and Hopt(k) represents the optimal vegetation height adjustment vector at observation point k.

7. The method for optimizing ecological landscaping design according to claim 6, characterized in that: S5 includes S51 and S52; S51. Combine the vegetation optimization density vector Pp, the visual obstruction matrix Um and the optimal vegetation height adjustment vector Hopt to calculate the comprehensive visual optimization index Q, which reflects the visual unobstructedness after the improvement of the landscape greening. The overall field of view optimization index Q is obtained through the following formula: ； In the formula, Mp represents the total amount of vegetation, Mu represents the total number of observation points, and U(k,j) represents the viewing obstacle index of vegetation j at observation point k. This represents the mean value of vegetation obstruction. Indicates the degree of deviation in vegetation height. The density reduction term is represented by q1, q2, and q3, which represent the weighting coefficients of the average vegetation obstruction, vegetation height deviation, and density reduction term, respectively, and q1+q2+q3=1. The specific values ​​are set by the user.

8. The method for optimizing ecological landscaping design according to claim 7, characterized in that: S52. Compare the obtained comprehensive vision optimization index Q with the preset optimization threshold Qth, obtain the comparison result, and determine whether to proceed to iterative optimization and decision execution based on the comparison result; The comparison results were obtained through the following comparison methods: When the comprehensive field of view optimization index Q ≤ optimization threshold Qth, it means that the comparison result is passed and the iterative optimization stage is not entered. Instead, the decision execution stage is entered, which includes generating vegetation height adjustment suggestions based on the optimal vegetation height adjustment vector Hopt, generating vegetation density adjustment suggestions based on the vegetation optimization density vector Pp, and generating occlusion observation point intervention suggestions based on the field of view obstacle matrix Um. When the comprehensive vision optimization index Q > the optimization threshold Qth, it indicates that the comparison result is not passed, and the system will not enter the decision execution environment, but will enter the iterative optimization stage, including executing S2 to S4 for iterative adjustment.

Images