An Automatic Identification Method for Urban Land Use Functions Based on Spatial Entropy

Abstract

This invention provides an automatic identification method for urban land use functions based on spatial entropy, relating to the field of urban planning. This method includes collecting geospatial information data of the target plot and case plot areas; vectorizing the geospatial information data; inputting the vectorized geospatial information data into a geographic information platform; calculating spatial entropy based on the geospatial information data of the case plot area; clustering the spatial entropy values ​​of the case plot area to form a spatial entropy value feature threshold library; delineating the land use boundaries of the target plot and identifying the plot function by comparing it with the spatial entropy value feature threshold library; calculating the window-to-wall ratio based on the street view data of the target plot to verify the land use function; outputting the target plot model information to a holographic screen for display; allowing planners to browse the model and select functions using gestures; and outputting the results as a land use map.

Description

Technical Field

[0001] This invention relates to the field of urban planning technology, specifically to an automatic identification method for urban land use functions based on spatial entropy. Background Technology

[0002] Accurately identifying the current land use functions of a city is a prerequisite for understanding its development, identifying its spatial structure, and conducting urban planning. Traditional land use identification requires manual visits and surveys of land use in various urban areas, followed by computer mapping and integration of the findings. This process is time-consuming and labor-intensive, making it unsuitable for studying medium to large-scale urban plots exceeding 50 hectares. Summary of the Invention

[0003] (a) Technical problems to be solved

[0004] To address the shortcomings of existing technologies, this invention provides an automatic identification method for urban land use functions based on spatial entropy. This method solves the problem that traditional land use identification requires manual visits and surveys of land use in various urban plots, followed by the mapping and integration of the survey data on a computer.

[0005] (II) Technical Solution

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] On the one hand, a method for automatic identification of urban land use functions based on spatial entropy is proposed, including:

[0008] Collect geospatial information data of the target plot and case plot, vectorize the geospatial information data, and input the vectorized geospatial information data into the geographic information platform;

[0009] Calculate spatial entropy based on geospatial information data of the case plot area;

[0010] Cluster the spatial entropy values ​​of the case plots to form a spatial entropy value feature threshold library;

[0011] Delineate the land use boundaries of the target plots and identify the functions of the plots by referring to the spatial entropy value feature threshold library;

[0012] The window-to-wall ratio is calculated based on the street view data of the target plot to verify the land use function;

[0013] The target plot model information is output to a holographic screen for display. Planners can browse the model and select functions based on gestures, and the results are output as a land use map.

[0014] Preferably, the process of collecting geospatial information data of the target plot and case plot ranges, vectorizing the geospatial information data, and inputting the vectorized geospatial information data into the geographic information platform specifically includes:

[0015] The geospatial information data collected for the target plot and case plot range includes road, building, and street view data. Road and building shape data are identified based on the spectral characteristics of remote sensing satellite images. Building height data is collected using drones equipped with image control point recognition devices. Street view data is obtained through the Baidu Maps street view platform. The remote sensing satellite images are obtained through the China Resources Satellite Data Service Network or public websites of the Geospatial Data Cloud.

[0016] The collected geospatial information data is vectorized and input into the geographic information platform. The land use patches and land use function data of the case plot are input, and the geospatial information data of the case plot is linked with the land use patches.

[0017] Preferably, the spatial entropy includes staggered entropy, area entropy, volume entropy, orientation entropy, and shape entropy.

[0018] Preferably, the staggered entropy represents the richness of the variation in building height of the plot, and is expressed by the formula O1 = m1 / H; where O1 is the staggered entropy, and m1 is the standard deviation of building height, expressed by the formula m1 = [(h1-H)]. 2 +(h2-H) 2 +……+(hn-H) 2 H is the average height, and the formula is H = (h1 + h2 + ... + hn) / n, where h1, h2, ..., hn are the average heights. n The height of each building within the plot;

[0019] Area entropy O2 represents the richness of variation in the building footprint of a plot, and is expressed by the formula O2 = m2 / A; where O2 is the area entropy, and m2 is the standard deviation of the building footprint, expressed by the formula m2 = [(a1-A)]. 2 +(a²-A) 2 +……+(a n -A) 2 ] / n, where A is the average land area, and the formula is A=(a1+a2+……+a n ) / n, a1、a2……a n The footprint of each building within the plot is obtained using computational geometry commands in the geographic information platform.

[0020] The volume entropy O3 represents the richness of the variation in building volume of the plot, and the formula is O3 = m3 / V; where O3 is the volume entropy, and m3 is the standard deviation of building area, and the formula is m3 = [(v1-V)]. 2 +(v2-V)2 +……+(v n -V) 2 ] / n, where V is the average area, and the formula is V=(v1+v2+……+v n ) / n,v1、v2……v n The volume of each building within the plot is given by the formula v. n =h n *a n ;

[0021] The orientation entropy O4 represents the richness of the variation in building orientation on the plot, and is expressed by the formula O4 = m4 / R; where O4 is the orientation entropy, and m4 is the standard deviation of the building orientation angle, expressed by the formula m4 = [(r1-R)]. 2 +(r2-R) 2 +……+(r n -R) 2 ] / n, R is the average orientation angle of the building, and the formula is R=(r1+r2+……+r n ) / n r1, r2...r n The orientation of each building within the plot is determined by the minimum boundary geometry command in the geographic information platform, and the orientation of the building is determined by the direction angle of the short side of the rectangle.

[0022] The shape entropy O5 represents the richness of the variation in building shape on the plot, and the formula is O5 = m5 / T; where O5 is the shape entropy, and m5 is the aspect ratio of the building's circumscribed rectangle, and the formula is m5 = [(t1-T)]. 2 +(t2-T) 2 +……+(t n -T) 2 ] / n, where T is the average aspect ratio of the building's circumscribed rectangle, and the formula is T=(t1+t2+……+t n ) / n,t1、t2……t n This refers to the aspect ratio of the bounding rectangles of each building within the plot area; the bounding rectangles of the buildings are obtained through the minimum boundary geometry command in the geographic information platform.

[0023] By summing up the staggered entropy, area entropy, volume entropy, orientation entropy, and shape entropy of each plot, a spatial entropy set z is formed for any plot n. n =(O 1n O 2n O 3n O 4n ).

[0024] Preferably, the step of clustering the spatial entropy values ​​of the case plot range to form a spatial entropy value feature threshold library specifically includes:

[0025] Based on urban land use classification and planning and construction land standards, four major categories of land use function tags and 24 subcategories of land use function tags are extracted. The four major categories of land use function tags are: A) Public management and public service facilities land, B) Commercial and service facilities land, R) Residential land, M) Industrial land, and their subcategories. The subcategories of land use function tags are linked to the spatial entropy set z of the land parcels. n This results in a dataset Z containing 24 different land use categories. n The K-means algorithm was used to cluster the spatial entropy values ​​of the four major land use functions into 30 classes, forming a total of 120 subclass datasets K1, K2...K 120 ;

[0026] Construct a KNN maximum nearest neighbor model, and use the small class dataset K n Features and the middle class dataset Z n Feature comparison was performed to extract the most similar land use class functions, and 120 subclass datasets were matched to 24 land use class function labels;

[0027] Compare the proportion of sets within the subclasses in each land use category, and remove subclasses with fewer than one-tenth of the average number of sets; assuming the number of sets in the subclasses within each land use category is x1, x2, x3...x n If any x j <x 平均 If x = 10, then the subclass dataset will be deleted, where x 平均 The formula for the average value is x. 平均 = (x1+x2+x3+……+x) n ) / n;

[0028] Summarize the feature values ​​of class F in a certain land use, and take the dataset K of each subclass. j The minimum and maximum values ​​of each element in the set (O1(K) j ) min -O1(K j ) max O2(K j ) min -O2(K j ) max ,O3(K j ) min -O3(K j ) max ,O3(K j ) min -O3(K j ) max Let j = 1, 2, 3...n, and form the characteristic thresholds P(F)1, P(F)2, P(F)3...P(F) for land use class F. n Furthermore, the feature thresholds of each land use category are summarized to form a spatial entropy value feature threshold library.

[0029] Preferably, the function of delineating the land use boundaries of the target plot and identifying the plot by referring to the spatial entropy value feature threshold library specifically includes:

[0030] The target plots are divided according to the main and secondary roads, and the spatial entropy value of the target plots is calculated in the manner described in S2. Among them, rectangles with a width greater than 30m are extracted from the road surface area of ​​the main and secondary roads.

[0031] The spatial entropy value of each target plot is compared with the feature threshold library in turn, and the land use function label of the corresponding threshold range is linked with the target plot.

[0032] Preferably, the step of calculating the window-to-wall ratio based on the street view data of the target plot to verify the land use function specifically includes:

[0033] The convolutional neural network is used to identify street scene features to form wall and window areas, and the building window-to-wall ratio is calculated using the formula E. 窗 / E 墙 E 窗 and E 墙 These are the areas of the walls and windows, respectively.

[0034] The window-to-wall ratio and land use function of the building were compared and selected according to Table 4. Plots that did not meet the window-to-wall ratio range were marked in red.

[0035] Preferably, the step of outputting the target plot model information to a holographic screen for display, allowing planners to browse the model and select functions based on gestures, and outputting the results as a land use map specifically includes:

[0036] The target plot model is output to a holographic digital sand table for three-dimensional display. The target plot model includes the target plot's spatial geographic information, building height model, land use function information, the range of the red-marked plot, and street view data.

[0037] The operator uses gesture commands to browse the model and select the land use function of the red-marked plots; the adjusted plot function labels are converted into layer color information and exported as a .dwg format land use map, with the color conversion of each function referring to the urban and rural planning land use classification and color block standards of various cities.

[0038] In another aspect, an apparatus is provided, the apparatus comprising:

[0039] One or more processors;

[0040] Memory, used to store one or more programs.

[0041] When the one or more programs are executed by the one or more processors, the one or more processors execute the urban land use function automatic identification method.

[0042] In another aspect, a computer-readable storage medium is provided that stores a computer program, which, when executed by a processor, implements the aforementioned method for automatic identification of urban land use functions.

[0043] (III) Beneficial Effects

[0044] 1. This invention provides an automatic identification method for urban land use functions based on spatial entropy, which reduces the time cost of land use function identification. Traditional identification methods require 2-4 weeks from land survey to manual processing to function judgment and data entry. This invention achieves automatic and intelligent identification of land use functions. The entire land use data collection, identification and verification process only takes 1-2 days, and the identification speed is increased by 14 times.

[0045] 2. The present invention provides an automatic identification method for urban land use functions based on spatial entropy, which reduces the manual cost of land use identification. Traditional identification methods require a team of 15 people to conduct functional surveys and verifications of land use of 10-30 hectares, and are difficult to apply to large and medium-scale cities of more than 50 hectares. The present invention uses digital and intelligent means, and only 1-2 people are needed to complete the functional identification of large and medium-scale cities.

[0046] 3. This invention provides an automatic identification method for urban land use functions based on spatial entropy. This expands the utilization of land use data, accurately expressing the spatial characteristics of land parcels using spatial entropy, thus improving the accuracy of function identification. Traditional methods rely on business type POI and land parcel construction indicator data, but data lag and incomplete coverage lead to low identification accuracy. This invention improves identification accuracy by 60%.

[0047] 4. The present invention provides an automatic identification method for urban land use functions based on spatial entropy, which realizes real-time verification, display and selection of function identification results. Planners can browse and operate the three-dimensional model of the land parcel through action commands, and further correct the land use functions based on morphology and street view data, thereby improving the efficiency of real-time interactive adjustment by planners by 21 times. Attached Figure Description

[0048] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0049] Figure 1 This is a flowchart illustrating the overall method and equipment framework of an embodiment of the present invention.

[0050] Figure 2 This is a data acquisition diagram according to an embodiment of the present invention;

[0051] Figure 3 This is a schematic diagram illustrating the spatial entropy calculation according to an embodiment of the present invention;

[0052] Figure 4 This is a diagram showing the automatic identification results of land use functions according to an embodiment of the present invention;

[0053] Figure 5 This is a schematic diagram illustrating the street view data verification in an embodiment of the present invention. Detailed Implementation

[0054] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0055] Example

[0056] like Figure 1 As shown in the figure, this invention discloses an automatic identification method for urban land use functions based on spatial entropy. The following urban land use identification of a certain area is used as an example and the accompanying drawings to illustrate the technical solution of this invention in detail, including the following steps:

[0057] S1. Collect geospatial information data of the target plot and case plot range, vectorize the above data and input it into the geographic information platform;

[0058] S2. Calculate the spatial entropy based on the geospatial information data of the case plot. The spatial entropy includes staggered entropy, area entropy, volume entropy, orientation entropy and shape entropy.

[0059] S3. Cluster the spatial entropy values ​​of the case plots to form a spatial entropy value feature threshold library;

[0060] S4. Delineate the land use boundaries of the target plot and identify the plot by referring to the spatial entropy value feature threshold library;

[0061] S5. Calculate the window-to-wall ratio based on the street view data of the target plot to verify the land use function;

[0062] S6. Output the target plot model information to the holographic screen for display. Planners can browse the model and select functions based on gestures, and the results are output as a land use map.

[0063] The above S1 includes two steps: S1.1 and S1.2.

[0064] S1.1 The geospatial information data of the collected land parcel includes road, building, and street view data. Road and building shape data are identified based on the spectral characteristics of remote sensing satellite imagery. Building height data is collected using a drone equipped with a ground truth (GST) recognition device. Street view data is obtained from Baidu Maps' street view platform. The remote sensing satellite imagery is obtained from the publicly available websites of the China Resources Satellite Data Service Network or the Geospatial Data Cloud.

[0065] S1.2. Vectorize the above data and input it into the geographic information platform, such as... Figure 2 As shown. Input the land use patches and land use function data of the case plot, and link the geospatial information data of the case plot with the land use patches;

[0066] 1. The above S2 includes seven steps: S2.1, S2.2, S2.3, S2.4, S2.5, S2.6, and S2.7.

[0067] S2.1 Calculate the spatial entropy based on the geospatial information data of the case. The spatial entropy includes staggered entropy, area entropy, volume entropy, orientation entropy and shape entropy, as shown in Table 1.

[0068] S2.2, The staggered entropy (O1) represents the richness of the variation in building height of a plot, and the formula is O1 = m1 / H; where m1 is the standard deviation of building height, and the formula is m1 = [(h1-H)]. 2 +(h2-H) 2 +……+(hn-H) 2 H is the average height, and the formula is H = (h1 + h2 + ... + hn) / n, where h1, h2, ..., hn are the average heights. n The height of each building within the plot.

[0069] S2.3, Area Entropy (O2) represents the richness of the variation in the building footprint of a plot, and the formula is O2 = m2 / A; where m2 is the standard deviation of the building footprint, and the formula is m2 = [(a1-A)]. 2 +(a²-A) 2 +……+(a n -A) 2 ] / n, where A is the average land area, and the formula is A=(a1+a2+……+a n ) / n, a1、a2……a n The footprint of each building within the plot is obtained using computational geometry commands in the geographic information platform.

[0070] S2.4, Volume Entropy (O3) represents the richness of the variation in building volume of a plot, with the formula O3 = m3 / V; where m3 is the standard deviation of building area, with the formula m3 = [(v1-V)]. 2 +(v2-V)2 +……+(v n -V) 2 ] / n, where V is the average area, and the formula is V=(v1+v2+……+v n ) / n,v1、v2……v n The volume of each building within the plot is given by the formula v. n =h n *a n .

[0071] S2.5, Orientation Entropy (O4) represents the richness of the variation in building orientation on a plot of land, and the formula is O4 = m4 / R; where m4 is the standard deviation of the building orientation angle, and the formula is m4 = [(r1-R)]. 2 +(r2-R) 2 +……+(r n -R) 2 ] / n, R is the average orientation angle of the building, and the formula is R=(r1+r2+……+r n ) / n r1, r2...r n The orientation of each building within the plot is determined by using the minimum boundary geometry command in the geographic information platform to obtain the minimum bounding rectangle of the building, and then using the direction angle of the short side of the rectangle as the building's orientation.

[0072] S2.6, Shape Entropy (O5) represents the richness of the variation in building shape on a plot of land, with the formula O5 = m5 / T; where m5 is the aspect ratio of the building's circumscribed rectangle, with the formula m5 = [(t1-T)]. 2 +(t2-T) 2 +……+(t n -T) 2 ] / n, where T is the average aspect ratio of the building's circumscribed rectangle, and the formula is T=(t1+t2+……+t n ) / n,t1、t2……t n This represents the aspect ratio of the bounding rectangles of all buildings within the plot. The building bounding rectangles are obtained using the minimum boundary geometry command in the geographic information platform.

[0073] S2.7. Summarize the staggered entropy, area entropy, volume entropy, orientation entropy, and shape entropy of each plot to form the spatial entropy set z for any plot n. n =(O 1n O 2n O 3n O 4n ).

[0074] Table 1: Types and Calculation Formulas of Spatial Entropy

[0075]

[0076]

[0077] The above S3 includes two steps: S3.1 and S3.2.

[0078] S3.1. Based on the Urban Land Use Classification and Planning Construction Land Use Standard (GB50137-2011), extract 4 major categories and 24 subcategories of land use function tags. The four major categories of land use function tags are: A. Public Management and Public Service Facilities Land; B. Commercial and Service Facilities Land; R. Residential Land; M. Industrial Land and subcategories thereof. Links to the subcategories of land use function tags and the spatial entropy set z of land parcels are provided. n This results in a dataset Z containing 24 different land use categories. n The K-means algorithm was used to cluster the spatial entropy values ​​of the four major land use functions into 30 classes, forming a total of 120 subclass datasets K1, K2...K. 120 As shown in Table 2;

[0079] Table 2: Minor Class Datasets

[0080]

[0081]

[0082] S3.2 Construct the KNN maximum nearest neighbor model, and use the small class dataset K n Features and the middle class dataset Z n Feature comparison was performed to extract the most similar land use class functions, and 120 subclass datasets were matched to 24 land use class function labels;

[0083] S3.3. Compare the proportion of sets in the sub-categories within each middle category dataset, and remove sub-categories whose sets are less than one-tenth of the average number. Assume the number of sets in the sub-categories within each land use middle category dataset is x1, x2, x3…x n If any x j <x 平均 If x = 10, then the subclass dataset will be deleted, where x 平均 The formula for the average value is x. 平均 = (x1+x2+x3+……+x) n ) / n;

[0084] S3.4 Summarize the feature values ​​of class F in a certain land use, and take the dataset K of each subclass. j The minimum and maximum values ​​of each element in the set (O1(K) j ) min -O1(K j ) max O2(K j ) min -O2(Kj ) max ,O3(K j ) min -O3(K j ) max ,O3(K j ) min -O3(K j ) max Let j = 1, 2, 3...n, and form the characteristic thresholds P(F)1, P(F)2, P(F)3...P(F) for land use class F. n Furthermore, the feature thresholds of each land use category are summarized to form a spatial entropy value feature threshold library, as shown in Table 3.

[0085] Table 3: Spatial Entropy Value Feature Threshold Library

[0086]

[0087]

[0088] The above S4 includes two steps: S4.1 and S4.2.

[0089] S4.1. Target land parcels are formed based on primary and secondary roads. The spatial entropy value of the target land parcels is calculated using the method described in S2. Figure 3 As shown, the main and secondary roads extract rectangles with a width greater than 30m from the road surface area;

[0090] S4.2. The spatial entropy values ​​of each target plot are compared sequentially with the feature threshold library. The land use function tags corresponding to the threshold ranges are linked to the target plots to obtain preliminary identification results of the plot functions. For example... Figure 4 As shown;

[0091] The above S5 includes two steps: S5.1 and S5.2.

[0092] S5.1 uses a convolutional neural network to identify street scene feature elements, forming wall and window areas, and calculates the building window-to-wall ratio using the formula E. 窗 / E 墙 E 窗 and E 墙 These are the areas of the walls and windows, respectively.

[0093] S5.2 Compare and select plots of land according to Table 4 based on the window-to-wall ratio and land use function, and mark plots that do not meet the window-to-wall ratio range in red, such as... Figure 5 As shown;

[0094] Table 4: Comparison Table of Land Use Functions and Building Window-to-Wall Ratio

[0095]

[0096] The above S6 includes two steps: S6.1 and S6.2.

[0097] S6.1 Output the target plot model to the holographic digital sand table for three-dimensional display. The target plot model includes the target plot's spatial geographic information, building height model, land use function information, the range of the red-marked plot, and street view data.

[0098] S6.2 The operator uses gesture commands to browse the model and select the land use function of the highlighted plots. The adjusted plot function labels are converted into layer color information and exported as a .dwg format land use map. The color conversion of each function is based on the urban and rural planning land use classification and color block standards of various cities.

[0099] As another embodiment of the present invention, a device is provided, the device comprising:

[0100] One or more processors;

[0101] Memory, used to store one or more programs.

[0102] When the one or more programs are executed by the one or more processors, the one or more processors execute the urban land use function automatic identification method in the above embodiments.

[0103] As another embodiment of the present invention, a computer-readable storage medium storing a computer program is provided, which, when executed by a processor, implements the urban land use function automatic identification method described in the above embodiments.

[0104] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

Claims

1. A method for automatic identification of urban land use functions based on spatial entropy, characterized in that, include: Collect geospatial information data of the target plot and case plot, vectorize the geospatial information data, and input the vectorized geospatial information data into the geographic information platform; Calculate spatial entropy based on geospatial information data of the case plot area; Cluster the spatial entropy values ​​of the case plots to form a spatial entropy value feature threshold library; Delineate the land use boundaries of the target plots and identify the functions of the plots by referring to the spatial entropy value feature threshold library; The window-to-wall ratio is calculated based on the street view data of the target plot to verify the land use function; The target plot model information is output to a holographic screen for display. Planners can browse the model and select functions based on gestures, and the results are output as a land use map. The spatial entropy includes staggered entropy, area entropy, volume entropy, orientation entropy, and shape entropy; The staggered entropy represents the richness of the variation in building height on a plot of land, and is expressed by the formula O1 = m1 / H; where O1 is the staggered entropy, and m1 is the standard deviation of building height, expressed by the formula m1 = [(h1-H)]. 2 +(h2-H) 2 +……+(hn-H) 2 H is the average height, and the formula is H = (h1 + h2 + ... + hn) / n, where h1, h2, ..., hn are the average heights. n The height of each building within the plot; Area entropy O2 represents the richness of variation in the building footprint of a plot, and the formula is O2=m2 / A; where O2 is the area entropy, and m2 is the standard deviation of the building footprint, and the formula is m2=[(a1-A)]. 2 +(a2-A) 2 +……+(a n -A) 2 ] / n, where A is the average land area, and the formula is A=(a1+a2+……+a n ) / n , a1、a2……a n The footprint of each building within the plot is obtained using computational geometry commands in the geographic information platform. The volume entropy O3 represents the richness of the variation in building volume of the plot, and the formula is O3=m3 / V; where O3 is the volume entropy, and m3 is the standard deviation of building area, and the formula is m3=[(v1-V)]. 2 +(v2-V) 2 +……+(v n -V) 2 ] / n, where V is the average area, and the formula is V=(v1+v2+……+v n ) / n , v1、v2……v n The volume of each building within the plot is given by the formula v. n =h n *a n ; The orientation entropy O4 represents the richness of the building orientation variation on the plot, and the formula is O4=m4 / R; where O4 is the orientation entropy, and m4 is the standard deviation of the building orientation angle, and the formula is m4=[(r1-R)]. 2 +(r2-R) 2 +……+(r n -R) 2 ] / n, R is the average orientation angle of the building, and the formula is R=(r1+r2+……+r n ) / n r1、r2……r n The orientation of each building within the plot is determined by the minimum boundary geometry command in the geographic information platform, and the orientation of the building is determined by the direction angle of the short side of the rectangle. The shape entropy O5 represents the richness of the variation in building shape on the plot, and the formula is O5=m5 / T; where O5 is the shape entropy, and m5 is the aspect ratio of the building's circumscribed rectangle, and the formula is m5=[(t1-T)]. 2 +(t2-T) 2 +……+(t n -T) 2 ] / n, where T is the average aspect ratio of the building's circumscribed rectangle, and the formula is T=(t1+t2+……+t n ) / n , t1、t2……t n This refers to the aspect ratio of the bounding rectangles of each building within the plot area; the bounding rectangles of the buildings are obtained through the minimum boundary geometry command in the geographic information platform. By summing up the staggered entropy, area entropy, volume entropy, orientation entropy, and shape entropy of each plot, a spatial entropy set z is formed for any plot n. n =(O 1n O 2n O 3n O 4n ).

2. The method for automatic identification of urban land use functions based on spatial entropy according to claim 1, characterized in that: The process of collecting geospatial information data of the target plot and case plot ranges, vectorizing the geospatial information data, and inputting the vectorized geospatial information data into the geographic information platform specifically includes: The geospatial information data collected for the target plot and case plot range includes road, building, and street view data. Road and building shape data are identified based on the spectral characteristics of remote sensing satellite images. Building height data is collected using drones equipped with image control point recognition devices. Street view data is obtained through the Baidu Maps street view platform. The remote sensing satellite images are obtained through the China Resources Satellite Data Service Network or public websites of the Geospatial Data Cloud. The collected geospatial information data is vectorized and input into the geographic information platform. The land use patches and land use function data of the case plot are input, and the geospatial information data of the case plot is linked with the land use patches.

3. The method for automatic identification of urban land use functions based on spatial entropy according to claim 1, characterized in that: The process of clustering the spatial entropy values ​​of the case plots to form a spatial entropy value feature threshold library specifically includes: Based on urban land use classification and planning and construction land standards, four major categories of land use function tags and 24 subcategories of land use function tags are extracted. The four major categories of land use function tags are: A) Public management and public service facilities land, B) Commercial and service facilities land, R) Residential land, M) Industrial land, and their subcategories. The subcategories of land use function tags are linked to the spatial entropy set z of the land parcels. n This results in a dataset Z containing 24 different land use categories. n The K-means algorithm was used to cluster the spatial entropy values ​​of the four major land use functions into 30 classes, forming a total of 120 subclass datasets K1, K2...K 120 ; Construct a KNN maximum nearest neighbor model, and use the small class dataset K n Features and the middle class dataset Z n Feature comparison was performed to extract the most similar land use class functions, and 120 subclass datasets were matched to 24 land use class function labels; Compare the proportion of sets within the subclasses in each land use category, and remove subclasses with fewer than one-tenth of the average number of sets; assuming the number of sets in the subclasses within each land use category is x1, x2, x3...x n If any x j <x 平均 If x = 10, then the subclass dataset will be deleted, where x 平均 The formula for the average value is x. 平均 = (x1 + x2 + x3 + ... + x n ) / n; Summarize the feature values ​​of class F in a certain land use, and take the dataset K of each subclass. j The minimum and maximum values ​​of each element in the set (O1(K) j ) min - O1(K j ) max O2(K) j ) min - O2(K j ) max O3(K) j ) min - O3(K j ) max O3(K) j ) min - O3(K j ) max Let j = 1, 2, 3...n, forming the characteristic thresholds P(F)1, P(F)2, P(F)3...P(F) for land use class F. n Furthermore, the feature thresholds of each land use category are summarized to form a spatial entropy value feature threshold library.

4. The method for automatic identification of urban land use functions based on spatial entropy according to claim 1, characterized in that: The function of delineating the land use boundaries of target plots and identifying plots by comparing them with a spatial entropy value feature threshold library specifically includes: The target plots are divided according to the main and secondary roads, and the spatial entropy value of the target plots is calculated in the manner described in S2. Among them, rectangles with a width greater than 30m are extracted from the road surface area of ​​the main and secondary roads. The spatial entropy value of each target plot is compared with the feature threshold library in turn, and the land use function label of the corresponding threshold range is linked with the target plot.

5. The method for automatic identification of urban land use functions based on spatial entropy according to claim 1, characterized in that: The step of calculating the window-to-wall ratio based on street view data of the target plot to verify the land use function specifically includes: The convolutional neural network is used to identify street scene features to form wall and window areas, and the building window-to-wall ratio is calculated using the formula E. 窗 / E 墙 E 窗 and E 墙 These are the areas of the walls and windows, respectively. The window-to-wall ratio and land use function of the building were compared and selected according to Table 4. Plots that did not meet the window-to-wall ratio range were marked in red.

6. The method for automatic identification of urban land use functions based on spatial entropy according to claim 1, characterized in that: The process of outputting the target plot model information to a holographic screen for display, allowing planners to browse the model and select functions using gestures, and outputting the results as a land use map, specifically includes: The target plot model is output to a holographic digital sand table for three-dimensional display. The target plot model includes the target plot's spatial geographic information, building height model, land use function information, the range of the red-marked plot, and street view data. The operator uses gesture commands to browse the model and select the land use function of the red-marked plots; the adjusted plot function labels are converted into layer color information and exported as a .dwg format land use map, with the color conversion of each function referring to the urban and rural planning land use classification and color block standards of various cities.

7. A device, characterized in that, The device includes: One or more processors; Memory, used to store one or more programs. When the one or more programs are executed by the one or more processors, the one or more processors perform the automatic identification method for urban land use functions as described in any one of claims 1-6.

8. A computer-readable storage medium storing a computer program, characterized in that, When executed by the processor, the program implements the automatic identification method for urban land use functions as described in any one of claims 1-6.

Images