A green space network optimization method and system for urban shrinkage

Abstract

The application discloses a kind of green space network optimization method and system for urban shrinkage, it is related to landscape architecture, urban and rural planning technical field, through high-resolution remote sensing image identification vacant land and combining machine learning algorithm establishes vacant land identification model, analysis vacancy rate and adopts K-means clustering division grid unit type, constructs vacant land conversion potential evaluation database and determines index weight using objective weighting method.Green space network is constructed based on graph theory model and the connection of node is analyzed using gravity model.Setting three target preference scenarios of life, ecology and production, and simulating network performance changes through global efficiency to determine the optimization threshold, finally combining network topology structure index and potential value, the importance and conversion priority of node are divided.It can scientifically evaluate the potential of green space in the process of urban shrinkage, optimize the spatial layout, improve the overall performance of green space network, and provide scientific support for the planning and management of green space.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of landscape architecture and urban and rural planning, in particular to a green space network optimization method and system for urban contraction. BACKGROUND

[0002] With the transformation of development patterns and the remodeling of urban functions in resource-exhausted cities and population-depleted cities, green space layout optimization has become an important strategy to cope with urban contraction, improve land use efficiency, and improve the ecological environment. In the context of urban contraction, the ecological, productive, and living functions of green space are particularly critical, and scientifically constructing a green space network can help alleviate the impact of reduced demand for public service facilities due to population reduction and enhance the resilience and sustainability of urban ecosystems. However, due to the complex distribution and frequent changes of vacant land in urban contraction areas, potential assessment and network optimization of green space layout face many challenges.

[0003] Currently, traditional methods for green space network optimization rely heavily on qualitative analysis and small-scale experimental research, usually through field surveys, historical data analysis, and other methods. These methods are limited by large resource inputs and poor timeliness, making it difficult to dynamically assess and systematically optimize a large range of vacant land of multiple types, and unable to fully reveal the complex relationship between green space layout and its ecological, productive, and living functions. SUMMARY

[0004] To solve the problems mentioned in the background, the purpose of the present application is to provide a green space network optimization method and system for urban contraction, which can assess the development potential of vacant space during urban contraction, optimize the layout of green space network, and provide support for the planning and management of green space.

[0005] In the first aspect, the purpose of the present application can be achieved by the following technical solution: a green space network optimization method for urban contraction, the method comprising the following steps:

[0006] Obtain high-resolution remote sensing images of vacant land, input the high-resolution remote sensing images of vacant land into a pre-trained vacant land recognition model, and output a green space development candidate land image;

[0007] Divide the green space development candidate land image into a grid to obtain a plurality of grid cells, obtain the vacancy rate of each grid cell, and combine K-means clustering to divide the grid cells into types and patterns, respectively, to obtain a green space supplement strategy framework;

[0008] The objective weighting method is used to calculate the weight of the pre-constructed vacant land conversion potential evaluation database to obtain a comprehensive weight value; the pre-constructed vacant land conversion potential evaluation database is constructed based on the dimensions of life, ecology and production.

[0009] The pre-set green space conversion scenarios with different target preferences are combined with the comprehensive weight value to obtain an input sequence, and the green space conversion scenarios with different target preferences are input into the pre-established green space network model based on the input sequence to obtain a network global optimization efficiency.

[0010] The importance of the node is calculated based on the network global optimization efficiency and the comprehensive weight value, the conversion priority is obtained based on the importance of the node, the type of green space suitable for supplementing in different grid units is determined according to the importance of the node and the conversion priority in combination with the green space supplement strategy framework, and the optimized urban green space network is generated based on the type of green space suitable for supplementing in different grid units.

[0011] In combination with the first aspect, in some implementations of the first aspect, the method further includes: the high-resolution remote sensing image of the vacant land relies on Google Earth to obtain high-definition images of the central urban area at preset intervals and cut them into tiles, and n tiles covering various types of urban vacant land and non-vacant land are uniformly selected for vacant land sample labeling.

[0012] In combination with the first aspect, in some implementations of the first aspect, the method further includes: the pre-trained vacant land recognition model uses a DeepLabv3 image segmentation model and Resnet34 as a backbone network for training, and precision, recall and F1 score are used as indicators for evaluating the recognition accuracy of the model.

[0013] The pre-trained vacant land recognition model is expressed as follows:

[0014] The calculation process of precision, recall and F1 is as follows:

[0015] Precision =

[0016] Recall =

[0017] F1 = 2x

[0018] In the formula, TP is the number of correctly predicted positive classes, FP is the number of incorrectly predicted positive classes, and FN is the number of incorrectly predicted negative classes.

[0019] The high-resolution remote sensing image of the vacant land is input into a pre-trained vacant land recognition model, and the vacancy time of different vacant lands is obtained through superimposed analysis, the permanent and temporary vacant lands are determined, the permanent vacant land is taken as the green space development candidate land, and finally the green space development candidate land image is obtained.

[0020] In combination with the first aspect, in some implementations of the first aspect, the method further includes that the calculation process of the vacancy rate of each grid cell is as follows:

[0021] R i =

[0022] In the formula, R i is the vacancy rate of the i-th grid cell, is the total area of the vacant land of the i-th grid cell, S i is the total area of the i-th grid cell.

[0023] In combination with the first aspect, in some implementations of the first aspect, the method further includes that the process of dividing the grid cells into types and modes by K-means clustering includes:

[0024] The grid cells are divided into four types of no vacancy, low vacancy, medium vacancy and high vacancy by K-means clustering;

[0025] The difference between the final value and the initial value of the vacancy rate is obtained at a preset interval, and the units are divided into three modes of contraction, stability and growth in combination with the overall change value;

[0026] Among them, the units of the contraction mode should be preferentially supplemented with production and ecological green space; the units of the stability mode are preferentially supplemented with life and ecological green space; and the units of the growth mode are preferentially supplemented with life green space.

[0027] In combination with the first aspect, in some implementations of the first aspect, the method further includes that the dimension indicators of the life, ecology and production include:

[0028] The life dimension indicators select four indicators of population density, park competition effect, adjacent facility support effect and slow walking network connection effect to respectively evaluate demand intensity, the influence of park competition on demand, facility support effect and the connection of the plot with the slow walking network; the ecological dimension indicators select four indicators of turning radius, fractal dimension index, patch importance index and connectivity comprehensive index to quantify the area characteristics, shape complexity, contribution to network connectivity and overall connectivity of the plot; the production dimension selects four indicators of slope, sunshine condition, transportation convenience and irrigation convenience to respectively analyze plot slope suitability, sunshine effective area proportion and distance from traffic roads and water systems;

[0029] Using the objective weighting method, the comprehensive weights of each indicator are generated to obtain the comprehensive potential value of each dimension, and the calculation process is as follows:

[0030]

[0031] In the formula: P j is the comprehensive potential value of the jth dimension; W j is the comprehensive weight of the jth indicator; I j is the standardized value of the jth indicator.

[0032] In combination with the first aspect, in some implementations of the first aspect, the method further comprises: the pre-set different target preference green space conversion scenarios include: life potential priority: connecting into the green space network from high to low according to the evaluation results of node life potential; ecological potential priority: connecting into the green space network from high to low according to the evaluation results of node ecological potential; production potential priority: connecting into the green space network from high to low according to the evaluation results of node production potential;

[0033] The construction process of the pre-established green space network model is as follows:

[0034] Select at least idle plots for a preset period of time, and simulate node connection based on existing parks to determine the best optimized network and its conversion timing;

[0035] Obtain city road data, including crawling open street map city road network data through Python's OSMnx package, adding three-level resistance points to model the time cost of connectivity, applying the travel OD cost matrix to analyze the time cost of node connectivity, using the gravity model to analyze the interaction between nodes, and creating node connection;

[0036] Using the city road network as the basis, the city current parks and idle land are simplified into network nodes, and the shortest path and time cost are used as edge weights to construct the green space network model;

[0037] The interaction between nodes is analyzed by a gravity model, and a connection length threshold is set according to the size of the node, wherein the connection rule between nodes is based on the area of two nodes and the distance relationship between them, and if the length threshold is exceeded, the node connection is not created.

[0038] The green space network model calculation process is as follows:

[0039] D(i,j) = Cost(O i ,D j )

[0040] In the formula, D(i,j) represents the time cost from the starting point O i to the ending point D j , and Cost(O i ,D j ) is a function of calculating cost.

[0041] In combination with the first aspect, in some implementations of the first aspect, the method further comprises: using a network global optimization efficiency index to measure the change of the overall performance of the network in the connection process, and relying on the network analysis library Network X of Python to program and realize the node connection simulation, to obtain the performance curve and the key optimization threshold of the network optimization, and the key optimization threshold is calculated as follows:

[0042] The calculation process of the network global optimization efficiency is as follows:

[0043] E(G) =

[0044] In the formula, E(G) is the network global optimization efficiency, n is the number of nodes in the network, i and j are any two nodes in the network, and d ij is the shortest cost path of node i and node j.

[0045] Taking the number of connected nodes as the abscissa and the network global optimization efficiency improvement rate as the ordinate, the performance change curve of the network in the connection process is fitted by a polynomial;

[0046] Using the zero first derivative and the first derivative lower than the average value as the key threshold of the performance improvement of the urban green space network.

[0047] In combination with the first aspect, in some implementations of the first aspect, the method further comprises: the importance of the node is obtained by equally weighting and superimposing the standardized node degree, betweenness degree, proximity degree and potential value of the node.

[0048] Superimpose the importance of each node in the three priority scenarios, divide the nodes into high, medium and low levels by K-means clustering, and correspond to high, medium and low conversion priorities, as follows:

[0049] S i =( + + + ) / 4

[0050] S i represent the importance of the node; D i , B i , P i , V i respectively represent the degree centrality weight, betweenness, proximity degree and potential value of node i, Max() is the maximum value, and Min() is the minimum value.

[0051] In a second aspect, to achieve the above object, the application discloses a green space network optimization system for urban contraction, comprising:

[0052] An image processing module is configured to obtain high-resolution remote sensing images of vacant land, input the high-resolution remote sensing images of vacant land into a pre-trained vacant land recognition model, and output green space development candidate land images.

[0053] A grid division module is configured to divide the green space development candidate land images into grid cells, obtain the vacancy rate of each grid cell, and combine K-means clustering to divide the grid cells into types and modes, thereby obtaining a green space supplement strategy framework.

[0054] A weight calculation module is configured to calculate the weights of a pre-constructed vacant land conversion potential evaluation database based on objective weighting method, and obtain a comprehensive weight value; wherein the pre-constructed vacant land conversion potential evaluation database is constructed based on the dimensions of life, ecology and production.

[0055] A model optimization module is configured to input different target preference green space conversion scenarios into a pre-established green space network model based on an input order obtained by combining the comprehensive weight value, and obtain a network global optimization efficiency.

[0056] The network output module is used for calculating the importance of the node based on the network global optimization efficiency and the comprehensive weight value, obtaining a conversion priority based on the importance of the node, determining a green space type suitable for supplementing in different grid units according to the importance of the node and the conversion priority in combination with a green space supplement strategy framework, and generating an optimized urban green space network based on the green space type suitable for supplementing in different grid units.

[0057] The present application has the following advantages:

[0058] The present application can evaluate the development potential of vacant space in the urban shrinkage process, optimize the layout of the green space network, and provide support for the planning and management of green space. BRIEF DESCRIPTION OF DRAWINGS

[0059] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art descriptions. Obviously, for those skilled in the art, other drawings can also be obtained based on these drawings without creative labor;

[0060] Figure 1 is a method flowchart of the present application;

[0061] Figure 2 is a green space supplement type schematic diagram of the present application;

[0062] Figure 3 is a green space network optimization schematic diagram of the present application;

[0063] Figure 4 is a system structure schematic diagram of the present application. DETAILED DESCRIPTION

[0064] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the protection scope of the present application.

[0065] Embodiment one:

[0066] As shown in Figure 1 A green space network optimization method for urban shrinkage, the method comprising the following steps:

[0067] obtaining a high-resolution remote sensing image of vacant land, inputting the high-resolution remote sensing image of vacant land into a pre-trained vacant land recognition model, and outputting a green space development candidate land image;

[0068] The high-resolution remote sensing images of vacant land rely on Google Earth to obtain high-definition images of the central urban area at five-year intervals and cut them into tiles. N tiles covering various types of urban vacant land and non-vacant land are uniformly selected for vacant land sample labeling.

[0069] Specifically, high-resolution remote sensing images of the central urban area of a certain city in 2014, 2019, and 2024 were collected using Google Earth, with a resolution of about 1.19 meters. After cutting the images into 2 km x 2 km tiles, 5 tiles covering different types of vacant land and non-vacant land were selected for labeling to enhance the model's recognition ability and robustness. After labeling, the images were cut into 224 x 224 pixel blocks. To further improve the accuracy of vacant land identification, we compared and analyzed three common image segmentation models: U-Net, FCN, and DeepLabv3. U-Net is suitable for small sample data and can provide good fine-grained segmentation results, but its segmentation accuracy is low in complex backgrounds. FCN is suitable for large-scale data processing, but it has limitations in detail segmentation and may require additional post-processing steps. DeepLabv3 can effectively capture large-scale context information through the hollow convolution mechanism, with high accuracy, especially in handling complex backgrounds and multi-scale features. Finally, DeepLabv3 combined with ResNet34 performed best in terms of precision, recall rate, and F1 score, with all three exceeding 80%. Therefore, we used the DeepLabv3 model and ResNet34 as the backbone network for model training and verification.

[0070] Table 1 Precision comparison of vacant land identification

[0071]

[0072] By training the qualified model on the central urban area images of a certain city in three years, we generated a vacant land distribution map and calculated the vacancy time of each land parcel through overlay analysis. The results showed that long-term vacant land (vacancy time ≥ 5 years) accounted for 7.17% of the total land area. These parcels had a low probability of being converted to other uses and were prioritized as potential green space optimization land.

[0073] The green space development candidate land image was divided into multiple grid units, the vacancy rate of each grid unit was obtained, and the grid units were classified by type and pattern using K-means clustering to obtain a green space supplement strategy framework.

[0074] The pre-trained vacant land identification model uses a DeepLabv3 image segmentation model, Resnet34 as a backbone network for training, and precision, recall and F1 as evaluation model identification accuracy indicators,

[0075] The pre-trained vacant land identification model is expressed as follows:

[0076] The calculation process of precision, recall and F1 is as follows:

[0077] Precision =

[0078] Recall =

[0079] F1 = 2x

[0080] In the formula, TP is the number of correctly predicted positive classes, FP is the number of incorrectly predicted positive classes, and FN is the number of incorrectly predicted negative classes.

[0081] The high-resolution remote sensing image of the vacant land is input into the pre-trained vacant land identification model, and the vacancy time of different vacant lands is obtained by superposition analysis, and the permanent and temporary vacant lands are determined, and the permanent vacant land is taken as the green space development candidate land, and finally the green space development candidate land image is obtained.

[0082] The calculation process of obtaining the vacancy rate of each grid cell is as follows:

[0083] R i =

[0084] In the formula, R i is the vacancy rate of the i-th grid cell, V i is the total area of the i-th grid cell, is the total area of the i-th grid cell.

[0085] The process of combining K-means clustering to divide the grid cells into types and modes includes:

[0086] The grid cells are divided into four types of no vacancy, low vacancy, medium vacancy and high vacancy by K-means clustering, and are respectively assigned values of 0, 1, 2 and 3;

[0087] The difference between the final value and the initial value of the vacancy rate is obtained by taking the preset year as an interval, and the overall change value is obtained, and the units are divided into three modes of shrinkage, stability and growth;

[0088] The production and ecological green space is preferentially supplemented in the unit of the shrinkage mode; the living and ecological green space is preferentially supplemented in the unit of the stable mode; and the living green space is preferentially supplemented in the unit of the growth mode.

[0089] Specifically, the research area is divided into 1km*1km grid units, and after calculating the vacancy rate, the grid is divided into "no vacancy", "low vacancy", "medium vacancy" and "high vacancy" four types through K-means clustering. Figure 2 The city green space supplement type schematic diagram of the present application is shown. The results show that the proportion of high vacancy grid increases from 0.34% to 7.22% in the past ten years, mainly distributed in the urban fringe area. By analyzing the vacancy rate change trend from 2014 to 2024, the grid units are divided into "shrinkage mode" (vacancy rate increases), "stable mode" (vacancy rate is basically unchanged) and "growth mode" (vacancy rate decreases), and the results show that the shrinkage mode unit of a certain central city accounts for as high as 57.03%, while the growth mode unit only accounts for 19.26%. Combined with the local development mode, the green space supplement strategy is set, wherein the production and ecological green space is preferentially supplemented in the shrinkage mode unit, the living and ecological green space is preferentially supplemented in the stable mode unit, and the living green space is preferentially supplemented in the growth mode unit, so as to adapt to the development needs of different regions. As shown in Table 2 below;

[0090] Table 2 Definition of local development mode of city

[0091]

[0092] An objective weighting method is used to calculate the weight of the pre-constructed vacancy land conversion potential evaluation database to obtain a comprehensive weight value; wherein the pre-constructed vacancy land conversion potential evaluation database is constructed based on the life, ecology and production dimension indexes;

[0093] The life dimension index, the ecology dimension index and the production dimension index are respectively selected.

[0094] As shown in Table 3 below, the life dimension index selects four indexes of population density, park competition effect, adjacent facility support effect and slow walking network connection effect to respectively evaluate the demand intensity, the influence of park competition on demand, the facility support effect and the connection of the plot with the slow walking network; the ecology dimension index selects four indexes of turning radius, fractal dimension index, patch importance index and connectivity comprehensive index to quantify the area characteristics, shape complexity, contribution to network connectivity and overall connectivity of the plot; and the production dimension selects four indexes of slope, sunshine condition, transportation convenience and irrigation convenience to respectively analyze the plot slope suitability, sunshine effective area proportion and distance from traffic road and water system;

[0095] To overcome the bias that may be caused by a single method, the CRITIC method and the entropy weight method are combined to generate a comprehensive weight value through linear weighted fusion. The CRITIC method assesses the contrast intensity between indexes by calculating the correlation between indexes, and the greater the weight of an index, the more independent the information quantity of the index in multi-dimensional data. The weight calculation formula is:

[0096]

[0097] In the formula, Critic j is the contrast intensity weight of the jth index; C j is the contrast intensity of the jth index.

[0098] The entropy weight method determines the weight based on the dispersion of each index, and the greater the dispersion of an index, the greater the weight. The weight formula is:

[0099]

[0100] In the formula, Entropy j is the dispersion weight of the jth index, E j is the contrast intensity of the jth index, which is calculated based on the correlation between the index and other indexes.

[0101] The CRITIC method weight and the entropy weight method weight are further combined to generate a comprehensive weight value by linearly weighting and fusing the contrast intensity weight calculated by the CRITIC method and the dispersion weight calculated by the entropy weight method. The calculation process is as follows:

[0102] W j = αEntropy j +(1−α)Critic j

[0103] In the formula, W j is the comprehensive weight of the jth index, α is a linear weighting coefficient, α ∈ [0, 1], and is used to balance the weight contribution of the CRITIC method and the entropy weight method.

[0104] Finally, the comprehensive potential value of each dimension is generated, and the calculation process is as follows:

[0105]

[0106] In the formula: P j is the comprehensive potential value of the jth dimension, W j is the comprehensive weight of the jth index, and I j is the standardized value of the jth index.

[0107] The Getis-Ord Gi* statistical method was used to analyze the spatial aggregation effect of the potential value, and the results showed that the aggregation effect of the ecological potential hotspots was weak as a whole, and a small number of hotspots were distributed in the center and northwest of the city, the production potential hotspots were mainly located in the northwest of the city, and the cold spots were distributed in the northeast; the living potential hotspots were mainly distributed in the central city, and the cold spots were scattered in the edge of the city.

[0108] Table 3 Construction of index system

[0109]

[0110] The pre-set green space conversion scenarios with different target preferences are converted, and the input order is obtained by combining the comprehensive weight value; different target preference green space conversion scenarios are input into the pre-established green space network model based on the input order, and the network global optimization efficiency is obtained;

[0111] The pre-set green space conversion scenarios with different target preferences include: living potential priority: connecting into the green space network from high to low according to the evaluation results of node living potential; ecological potential priority: connecting into the green space network from high to low according to the evaluation results of node ecological potential; production potential priority: connecting into the green space network from high to low according to the evaluation results of node production potential;

[0112] The construction process of the pre-established green space network model includes:

[0113] Selecting land parcels that have been idle for at least 5 years, and simulating node connection based on existing parks in 2024 to determine the best optimized network and its conversion timing;

[0114] The OSMnx package of Python is used to crawl the open street map city road network data, and three-level resistance points are added to model the time cost of connectivity, the OD cost matrix of travel is used to analyze the time cost of node connectivity, the gravity model is used to analyze the interaction between nodes, and node connection is created;

[0115] Using the city road network as the basis, the city current parks and idle land are simplified as network nodes, and the shortest path and time cost are used as edge weights to construct the green space network model;

[0116] The interaction between nodes is analyzed by the gravity model, and the connectivity length threshold is set according to the size of the node area, wherein the connection rule between nodes is based on the area and distance relationship between two nodes, and the node connection is not created if the length threshold is exceeded.

[0117] Specifically, the green space network is constructed based on graph theory method, simplifying existing parks and potential land into network nodes, taking the shortest path time cost as the edge weight, and adding fast road intersections (113s), main road intersections (63s), and secondary road intersections (43s) as three levels of resistance points for path optimization combined with road data extracted by OSMnx tool. The node connection rule sets a connection length threshold according to the size of the node area to ensure the scientificity and flexibility of network connection. The results show that the node distribution is relatively random in the ecological priority scenario, the nodes are concentrated in the edge area in the production priority scenario, and the nodes are concentrated in the central urban area in the life priority scenario.

[0118] The green space network model is as follows:

[0119] D(i,j) = Cost(O i ,D j )

[0120] In the formula, D(i,j) represents the time cost from the starting point O i to the ending point D j , and Cost(O i ,D j ) is a function for calculating the cost.

[0121] The study sets three scenarios of ecological potential priority, production potential priority, and life potential priority, respectively simulating the layout optimization process of green space network with priority of life, ecology, and production value. In each scenario, the nodes are connected into the network in order of high to low potential value, forming an optimized layout scheme with different target preferences. The ecological priority scenario maximizes the ecological function of urban green space, the production priority scenario improves the economic efficiency of land in the edge area, and the life priority scenario improves the recreational accessibility of residents in the central urban area.

[0122] The calculation process of the global optimization efficiency of the network is as follows:

[0123] E(G) =

[0124] In the formula, E(G) is the global optimization efficiency of the network, n is the number of nodes in the network, i and j are any two nodes in the network, and d ij is the shortest cost path between node i and node j.

[0125] The importance of the node is calculated based on the global optimization efficiency of the network and the comprehensive weight value, the conversion priority is obtained based on the importance of the node, the type of green space suitable for supplementing in different grid cells is determined based on the importance of the node and the conversion priority combined with the green space supplement strategy framework, and the optimized urban green space network is generated based on the type of green space suitable for supplementing in different grid cells.

[0126] Specifically, the performance change curve is fitted using the least square method, and the highest threshold of network performance is determined by taking the first derivative as zero, and the most cost-effective threshold is determined by taking the first derivative below the average value. Figure 3 As shown in the figure, the most cost-effective node numbers under different scenarios are 216 (ecological priority), 192 (production priority) and 312 (life priority), and the highest node numbers are 756 (ecological priority), 773 (production priority) and 819 (life priority).

[0127] The importance of the node is obtained by equal-weight superposition of the standardized node degree, betweenness, proximity and potential value of the node;

[0128] The importance of the node in the three priority scenarios is superimposed, and the nodes are divided into high, medium and low levels by K-means clustering, which correspond to high, medium and low conversion priorities respectively, as follows:

[0129] S i =( + + + ) / 4

[0130] S i represents the importance of the node; D i , B i , P i , V i respectively represent the weight, betweenness, proximity and potential value of the node i, Max() is the maximum value, and Min() is the minimum value.

[0131] Example 2: As shown in the figure, in order to achieve the above purpose, the application discloses a green space network optimization system for urban contraction, which comprises: Figure 4

[0132] The image processing module 11 is used for acquiring high-resolution remote sensing images of vacant land, inputting the high-resolution remote sensing images of vacant land into a pre-trained vacant land recognition model, and outputting to obtain a green space development candidate land image;

[0133] ​The grid division module 12 is configured to divide the green space development candidate land image into a plurality of grid cells, obtain the vacancy rate of each grid cell, and divide the grid cells into different types and modes by combining K-means clustering, so as to obtain a green space supplement strategy framework.

[0134] The weight calculation module 13 is configured to calculate the weight of a pre-constructed vacancy land conversion potential evaluation database by using an objective weighting method, so as to obtain a comprehensive weight value, wherein the pre-constructed vacancy land conversion potential evaluation database is constructed based on the dimensions of life, ecology and production.

[0135] The model optimization module 14 is configured to input a green space conversion scenario with different target preferences into a pre-established green space network model based on an input sequence obtained by combining the comprehensive weight value, so as to obtain a network global optimization efficiency.

[0136] The network output module 15 is configured to calculate the importance of a node based on the network global optimization efficiency and the comprehensive weight value, obtain a conversion priority based on the importance of the node, determine a suitable green space type to be supplemented in different grid cells based on the importance of the node and the conversion priority in combination with the green space supplement strategy framework, and generate an optimized urban green space network based on the suitable green space type to be supplemented in different grid cells.

[0137] Based on the same inventive concept, the application further provides a computer device, which comprises one or more processors and a memory for storing one or more computer programs; the program comprises program instructions, and the processor is configured to execute the program instructions stored in the memory. The processor can be a central processing unit (CPU), and can also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, and is configured to implement one or more instructions, and is specifically configured to load and execute one or more instructions in the computer storage medium to implement the above method.

[0138] It should be further noted that based on the same inventive concept, the present application further provides a computer storage medium, which stores a computer program, and the computer program is run by a processor to execute the above method. The storage medium can adopt any combination of one or more computer readable media. The computer readable medium can be a computer readable signal medium or a computer readable storage medium. The computer readable storage medium may, for example, but is not limited to, an electrical, magnetic, optical, electrical, magnetic, infrared, or semiconductor system, device or apparatus, or any combination thereof. More specific examples (non-exhaustive list) of the computer readable storage medium include: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, the computer readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, device or apparatus.

[0139] In the description of the present application, the description of the terms "one embodiment", "an example", "a specific example" and the like means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present disclosure. In the present specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0140] The above shows and describes the basic principles, main features and advantages of the present disclosure. It should be understood by those skilled in the art that the present disclosure is not limited by the above embodiments, and the above embodiments and descriptions in the specification are only to illustrate the principles of the present disclosure. Without departing from the spirit and scope of the present disclosure, various changes and improvements can be made to the present disclosure, and all these changes and improvements fall within the scope of the present disclosure.

Claims

1. A green space network optimization method for urban shrinkage, characterized in that, The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps:

2. The green space network optimization method for urban shrinkage according to claim 1, characterized in that, The method comprises the following steps:

3. The green space network optimization method for urban shrinkage according to claim 1, wherein, The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: Precision = Recall = F1 = 2x The method comprises the following steps: The method comprises the following steps:

4. The green space network optimization method for urban shrinkage according to claim 1, characterized in that, The method comprises the following steps: R i = In the formula, R i is the vacancy rate of the i-th grid cell, is the total area of vacant land of the i-th grid cell, is the total area of the i-th grid cell.

5. The green space network optimization method for urban shrinkage according to claim 1, wherein, The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the Among them, the unit of the contraction mode should be preferentially supplemented with production and ecological green space; the unit of the stable mode should be preferentially supplemented with life and ecological green space; the unit of the growth mode should be preferentially supplemented with life green space.

6. The green space network optimization method for urban shrinkage according to claim 1, wherein, The life, ecology and production dimension indicators respectively include: The life dimension indicators select four indicators of population density, park competition effect, adjacent facility support effect and slow walking network connection effect to respectively evaluate demand intensity, the influence of park competition on demand, facility support effect and the connection of the plot with the slow walking network; the ecological dimension indicators select four indicators of turning radius, fractal dimension index, patch importance index and connectivity comprehensive index to quantify the area characteristics, shape complexity, contribution to network connectivity and overall connectivity of the plot; the production dimension selects four indicators of slope, sunshine condition, transportation convenience and irrigation convenience to respectively analyze plot slope suitability, sunshine effective area proportion and distance from traffic roads and water systems; Then, the objective weighting method is used to generate the comprehensive weight value of each indicator to obtain the comprehensive potential value of each dimension, and the calculation process is as follows: In the formula, P j is the comprehensive potential value of the jth dimension, W j is the comprehensive weight of the jth index, I j is the normalized value of the jth index.

7. The green space network optimization method for urban shrinkage according to claim 6, characterized in that, The pre-set green space conversion scenarios with different target preferences include: life potential priority: connecting into the green space network from high to low according to the evaluation results of node life potential; ecological potential priority: connecting into the green space network from high to low according to the evaluation results of node ecological potential; production potential priority: connecting into the green space network from high to low according to the evaluation results of node production potential; The construction process of the pre-established green space network model is as follows: Select at least idle plots for a preset period of time, and simulate node connection based on existing parks to determine the optimal optimized network and its conversion timing; Obtain city road data, including crawling open street map city road network data through Python's OSMnx package, adding three-level resistance points to model the time cost of connectivity, applying the OD cost matrix of travel to analyze the time cost of node connectivity, using the gravity model to analyze the interaction between nodes, and creating node connection; wherein, the three-level resistance points include expressway intersections, main road intersections and secondary road intersections; Using the city road network as the basis, the city current park and idle land are simplified into network nodes, and the shortest path and time cost are used as edge weights to construct a green space network model; The interaction between nodes is analyzed by the gravity model, and the connection length threshold is set according to the size of the node area, wherein the connection rules between nodes are based on the area and distance relationship between two nodes, and if the set length threshold is exceeded, the node connection is not created; The green space network model calculation process is as follows: D(i,j) = Cost(O i ,D j ) where D(i,j) represents the time cost from the start point O i to the end point D j , and Cost(O i ,D j ) is a function of the computation cost.

8. The urban shrinkage-oriented green space network optimization method according to claim 7, characterized in that, The network global optimization efficiency index is used to measure the change of the overall performance of the network in the connection process, and the network analysis library Network X of Python is used for programming to realize node connection simulation, obtain the performance curve of network optimization and the key optimization threshold, and the calculation method of the key optimization threshold is as follows: The calculation process of the network global optimization efficiency is as follows: E(G) = E(G) + 1 where E(G) is the global optimization efficiency of the network, n is the number of nodes in the network, i and j are any two nodes in the network, d ij is the shortest cost path between node i and node j; Taking the number of connected nodes as the horizontal coordinate and the network global optimization efficiency improvement rate as the vertical coordinate, the performance change curve of the network during the connection process is fitted by a polynomial; Using the first derivative being zero and the first derivative being lower than the average value as the key threshold for performance improvement of urban green space network.

9. The green space network optimization method for urban shrinkage according to claim 8, characterized in that, The importance of the node is obtained based on the network global optimization efficiency and the comprehensive weight value, including: obtaining by equal-weight superposition after standardizing the node degree centrality Node degree, betweenness centrality Betweenness degree, proximity centrality Proximity degree and potential value; Superimposing the importance of each node in the three priority scenarios, the nodes are divided into high, medium and low levels by K-means clustering, which correspond to high, medium and low conversion priorities respectively, as follows: S i =( + + + ) / 4 S i denotes the node importance; D i , B i , P i , V i denote the weight, betweenness, proximity degree and potential value of node i respectively, Max() is the maximum value, and Min() is the minimum value.

10. A green space network optimization system for urban shrinkage, adopting the green space network optimization method for urban shrinkage according to any one of claims 1 to 9, characterized in that, It includes: The image processing module is used for acquiring high-resolution remote sensing images of vacant land, inputting the high-resolution remote sensing images of vacant land into a pre-trained vacant land recognition model, and outputting a green space development candidate land image; The grid division module is used for dividing the green space development candidate land image into grid cells, obtaining the vacancy rate of each grid cell, and combining K-means clustering to divide the grid cells into types and modes, thereby obtaining a green space supplement strategy framework; The weight calculation module is used for calculating the weight of the pre-constructed vacant land conversion potential evaluation database based on objective weighting method, and obtaining a comprehensive weight value; wherein the pre-constructed vacant land conversion potential evaluation database is constructed based on the dimensions of life, ecology and production; The model optimization module is used for inputting the pre-set green space conversion scenarios with different target preferences into the pre-established green space network model based on the input order obtained by combining the comprehensive weight value, and obtaining the network global optimization efficiency; The network output module is used for obtaining the importance of the node based on the network global optimization efficiency and the comprehensive weight value, obtaining the conversion priority based on the importance of the node, determining the suitable green space types to be supplemented in different grid cells according to the importance of the node and the conversion priority in combination with the green space supplement strategy framework, and generating an optimized urban green space network based on the suitable green space types to be supplemented in different grid cells.

Images