A method, medium and device for identifying urban difficult site spaces

Abstract

The present application relates to the technical field of quantitative identification of difficult sites, and particularly relates to a city difficult site space identification method, medium and equipment, wherein the city difficult sites are divided into two categories of natural type city difficult sites and artificial type city difficult sites according to different cause types and site types; the natural type city difficult sites and the artificial type city difficult sites are respectively integrated with existing land use classification data, the natural type city difficult sites are identified by using remote sensing images of the existing land use classification data, the artificial type city difficult sites are identified according to whether construction land transfer is needed as a standard combined with remote sensing images; and low coverage grasslands with NDVI less than 0.2 are supplemented as newly divided natural type city difficult sites; the identified natural type city difficult sites and artificial type city difficult sites are combined, and a complete city difficult site range is obtained. The present application can be specifically applied in city difficult site space identification in different scale ranges.

Description

Technical Field

[0001] This invention relates to the technical field of quantitative identification of difficult sites, specifically to a method, medium, and device for spatial identification of difficult sites in urban areas. Background Technology

[0002] Urban challenging sites refer to the general term for site spaces in urban areas that cannot meet the site conditions required for the normal growth of major zonal vegetation species. Existing research on urban challenging sites focuses on defining them, controlling pests and diseases in these sites, and selecting vegetation—studies based on understanding the nature of challenging sites. More in-depth research is lacking. This is because the identification of urban challenging sites currently relies more on on-site manual identification or accumulated experience, lacking a method for direct spatial identification. Zhang Lang's edited book, *Methods and Practices for Ecological Landscape Construction in Urban Challenging Sites*, divides urban challenging site types into three major categories and ten subcategories, including natural, degraded, and artificial types, comprehensively encompassing existing types of urban challenging sites. However, it struggles to quantitatively identify the spatial distribution range of urban challenging sites.

[0003] In addition, the current land use classification standard divides space into 12 primary categories and 73 secondary categories, including cultivated land, orchards, forest land, grassland, commercial and service land, industrial and mining storage land, residential land, public management and public service land, special land, transportation land, water area and water conservancy facilities land, and other land. However, there is no precise category for urban difficult sites, making it impossible to directly retrieve their scope from the spatial data.

[0004] Therefore, this invention proposes a new classification method that enables direct spatial identification of urban difficult site areas. Summary of the Invention

[0005] To address the shortcomings of existing definitions and classifications of urban difficult sites in terms of their inability to quantitatively delineate their spatial extent, this invention provides a method, medium, and device for spatial identification of urban difficult sites. This method combines the "Classification of Urban Land Use and Standards for Planning and Construction Land" (GB50137-2011), the "Land Administration Law of the People's Republic of China (2019 Revision)," and the definition of urban difficult sites. It also references existing classification methods for urban difficult sites and urban ecological land use to propose a new classification method for urban difficult sites that facilitates spatial identification. Based on this classification method, by utilizing existing land use data and remote sensing imagery data, and by analyzing and overlaying the two datasets and comparing land use data transfer patterns from adjacent years, the spatial extent of urban difficult sites can be determined.

[0006] The first aspect of this invention provides a method for identifying challenging urban sites, comprising the following steps:

[0007] Based on different causes and site types, urban difficult sites are divided into two main categories: natural urban difficult sites and artificial urban difficult sites.

[0008] The types of natural and artificial urban difficult sites are integrated with existing land use classification data. Natural urban difficult sites are identified using remote sensing images of existing land use classification data, while artificial urban difficult sites are identified by combining remote sensing images according to whether or not they need to be transferred for construction land.

[0009] Furthermore, by inverting NDVI using remote sensing image data, low-coverage grasslands with NDVI < 0.2 were added to the newly delineated natural-type urban difficult sites;

[0010] By merging the identified natural and artificial urban challenging sites, a complete range of urban challenging sites is obtained.

[0011] Natural-type challenging urban sites: Ecological land where site conditions are dominated by natural factors such as climate and geology, hindering plant growth. (Refer to *Methods and Practices for Ecological Landscape Construction on Challenging Urban Sites*, edited by Zhang Lang.)

[0012] Artificial urban challenging sites: Ecological land where site conditions are dominated by human interference factors such as engineering construction, land use type change, and pollutant emissions, which cause obstacles to plant growth or seriously damage the function of the ecosystem.

[0013] Transfer of construction land: Whether the land use type has been affected by human construction activities and has become construction land, and restoration is carried out on the basis of construction land. In this invention, it specifically refers to the situation where it is currently identified as green space but was previously construction land.

[0014] Preferably, natural urban challenging site types include saline-alkali land, grassland, sandy land, bare land and bare rock gravel land, glaciers and permanent snow cover, damaged wetlands or water bodies.

[0015] Preferably, the types of challenging sites in artificial urban areas include industrial relocation sites, controlled vacant land, landfills, and green spaces created by buildings.

[0016] Preferably, the existing land use classification data comes from the National Glacier, Permafrost and Desert Science Data Center. This data classifies land use into cultivated land, forest land, grassland, shrubland, construction land, unused land, water bodies, desert land, snow-capped mountain and glacier land and wetland. Unused land, desert land, snow-capped mountain and glacier land and wetland within the urban area are considered as natural-type urban difficult sites.

[0017] Preferably, the method for identifying natural-type urban difficult sites is as follows: the types of natural-type urban difficult sites are identified and overlaid using existing land use classification data, and then NDVI is retrieved from remote sensing image data. Low-coverage grasslands with NDVI < 0.2 are added as natural-type urban difficult sites. The two parts are overlaid to form the range of natural-type urban difficult sites.

[0018] Preferably, NDVI is retrieved from Landsat 8 remote sensing image data to obtain low-coverage grassland data with NDVI < 0.2.

[0019] In this invention, NDVI is retrieved using 30m Landsat 8 remote sensing image data extracted by the U.S. Geological Survey.

[0020] Preferably, the method for identifying artificially constructed urban difficult sites is as follows: Land use transfer data from two consecutive time periods are used for delineation. Plots currently identified as green space by remote sensing but previously used as construction land are identified as artificially constructed urban difficult sites. The land use data used in this method comes from the National Glacier, Permafrost and Desert Scientific Data Center. This data classifies land use into cultivated land, forest land, grassland, shrubland, construction land, unused land, water bodies, desert land, snow-capped mountain and glacier land, and wetlands. Cultivated land, forest land, grassland, and shrubland are reclassified as green space. Construction land and water bodies remain as construction land and water bodies. The remaining categories are classified as natural urban difficult sites. Finally, the nine land use types in the original data are reclassified into four categories: green space, construction land, natural urban difficult sites, and water bodies. The transfer trend and amount of construction land and green space are calculated using a land use transfer matrix. The portion transferred from construction land to green space is identified as artificially constructed difficult sites.

[0021] Land use transfer data: refers to data on land that is currently identified as green space but was previously construction land.

[0022] A second aspect of the present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the aforementioned method for spatial identification of difficult urban sites.

[0023] A third aspect of the present invention provides a non-transitory computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the aforementioned method for spatial identification of difficult urban sites.

[0024] A fourth aspect of the present invention provides a computer program product, including a computer program that, when executed by a processor, implements the urban difficult site spatial identification method.

[0025] The overall concept of this invention is as follows: Urban challenging sites, as an important component of urban spatial patterns and a significant source of urban green space, are of substantial importance to study. The lack of relevant research stems from the inability to spatially identify them, hindering further research. This invention proposes a classification method for urban challenging sites. Based on the definition of urban challenging sites and the current "Classification of Urban Land Use and Standards for Planning and Construction Land" (GB50137-2011) and the "Land Administration Law of the People's Republic of China (2019 Revision)," the types of urban challenging sites are determined. Furthermore, based on different causes and site types, urban challenging sites are divided into natural urban challenging sites and artificial urban challenging sites. This is integrated with existing land use classification data (the selected land use data comes from the National Glacier, Permafrost and Desert Scientific Data Center (http: / / www.ncdc.ac.cn)), allowing direct spatial identification of the two main categories of urban challenging sites: natural and artificial.

[0026] This invention classifies urban difficult sites into two major categories—natural and artificial—based on their formation and site type. It then integrates this classification with existing land use classification data (selected from the National Glacier, Permafrost and Desert Scientific Data Center (http: / / www.ncdc.ac.cn)). Natural and artificial urban difficult sites are matched into the nine categories of the existing land use classification data. This allows for direct identification and overlay of natural urban difficult sites based on remote sensing images in the existing database. Furthermore, by inverting NDVI from the remote sensing image data, low-coverage grasslands with NDVI < 0.2 are added as natural urban difficult sites. Artificial urban difficult sites are identified based on whether they require transfer for construction land, combined with remote sensing images. The identified natural and artificial urban difficult site ranges are then merged to obtain a complete range of urban difficult sites, enabling direct spatial identification of the urban difficult site range.

[0027] The present invention has the following advantages and beneficial effects:

[0028] This invention overcomes the limitation of existing concepts of urban difficult sites being difficult to use directly for quantitative identification of the spatial extent of urban difficult sites, and can be specifically applied in the spatial identification of urban difficult sites at different scales.

[0029] This invention proposes a new classification method for urban difficult sites, dividing them into two main categories—natural and artificial—based on their formation and site type. After integration with existing land use classification data, spatial identification can be directly performed, ensuring the comprehensiveness of the scope of urban difficult sites and the accuracy of spatial identification. This has important guiding significance for urban land use planning. Attached Figure Description

[0030] Figure 1 The image shows land use data extracted from the urban built-up area of ​​a city in 2020 in Example 2. The data is divided into four categories using ArcGIS: green space, unused land, construction land, and water bodies (left). The image also shows NDVI data obtained from vegetation inversion of the area using Landsat 8 remote sensing imagery (right).

[0031] Figure 2 This is a map showing the extent of naturally-type urban difficult sites in the built-up area of ​​a certain city in 2020, as shown in Example 2.

[0032] Figure 3 This is a schematic diagram illustrating the calculation of land use transfer data for 2020-2021 using the land use transfer matrix method in the intersection module of ArcGIS, based on the reclassified 2020 and 2021 data in Example 2.

[0033] Figure 4 This is a schematic diagram of the final urban hardship site area of ​​a certain city in 2020 in Example 2. Detailed Implementation

[0034] To better understand the present invention, the following embodiments are further illustrations of the present invention, but the content of the present invention is not limited to the following embodiments.

[0035] Example 1

[0036] Table 1 shows the current classification standards for challenging urban sites. The classification standards are derived from "Methods and Practices for Ecological Landscape Construction in Challenging Urban Sites" edited by Zhang Lang.

[0037] Table 1 Current Classification Standards for Urban Difficult Sites

[0038]

[0039]

[0040] The above classification divides urban difficult sites into three main categories: natural, degraded, and artificial. If these three categories are directly spatially identified, it may cause some overlap in identification. For example, the identification method for both the degraded type (town relocation sites) and the artificial type (sites converted from construction land to green space) is the same, and they will both be uniformly identified as construction land in space, which will lead to overlap between the two types.

[0041] To address the above problems, this invention proposes a novel method for identifying challenging urban sites, comprising the following steps:

[0042] The first step involved defining urban challenging sites based on their definition and causes, categorizing them into two main types: natural and artificial. Next, a comparison was made with the descriptions of land types in the *Classification of Urban Land Use and Standards for Planning and Construction Land* (GB 50137-2011) and the *Land Administration Law of the People's Republic of China (2019 Revision)*, and referenced existing research on urban ecological site types to determine the subcategories of urban challenging sites. Based on their formation and site type, urban challenging sites were further divided into two main categories: artificial and natural, as shown in Table 2.

[0043] Table 2. Optimized Classification of Urban Difficult Sites

[0044]

[0045] The second step involves selecting different identification methods based on the different causes of natural and artificial types:

[0046] Naturally challenging urban sites are distributed across the major categories of existing land use classification standards. This example utilizes existing land use classification data for identification and overlay. The land use data selected in this embodiment comes from the National Data Center for Glacier, Permafrost and Desert Science (http: / / www.ncdc.ac.cn). This dataset uses Landsat remote sensing imagery, providing 37 years of land cover type data for China. Training samples were collected through stable sample extraction and visual interpretation, and temporal indicators were constructed and obtained using a random forest classifier. The data quality is good. This data classifies land use into nine categories: cultivated land, forest land, grassland, shrubland, construction land, unused land, water bodies, desert land, snow-capped mountain and glacial land, and wetlands. Unused land, desert land, snow-capped mountain and glacial land, and wetlands within urban areas are considered naturally challenging urban sites.

[0047] Specifically, the definitions of types included in the natural type of difficult urban sites are compared with the definitions of the nine land use classifications in the dataset. Unused land refers to land other than agricultural land and construction land, mainly including grassland, saline-alkali land, swamp, sandy land, bare land, and bare rock. It includes two secondary land categories: unused land and other unused land. Unused land is further divided into: ① Grassland: Tree canopy closure <10%, topsoil, weeds growing, excluding saline-alkali land, swamp, and bare land. ② Saline-alkali land: Topsoil with salt accumulation, only naturally salt-tolerant plants growing. ③ Sandy land: Topsoil covered with sand, basically without vegetation, including deserts, excluding sandy beaches in water systems. ④ Bare land: Topsoil covered with soil, basically without vegetation. ⑤ Bare rock and gravel land: Topsoil covered with rocks or gravel, with a coverage area >50%. Other unused land is further divided into: ① Other land: Other water areas not included in agricultural land or construction land. ② River surface. ③ Land below the normal water level of the shoreline of naturally formed or artificially excavated rivers. ④ Lake surface. Land below the normal water level of the shoreline of naturally formed waterlogged areas. ⑤ Reed beds. Land where reeds grow, including reed beds on mudflats. ⑥ Glaciers and permanent snow cover. Land whose surface is covered by ice and snow year-round. The types of land use in the classification of difficult sites in natural cities have a high degree of overlap with those in the classification of difficult sites in natural cities. The sandy land and desert land types in the unused land are similar. However, the land use data used in this invention separately classifies desert land, and desert land also meets the classification of difficult sites in natural cities. Therefore, unused land, desert land, snow-capped mountains and glaciers, and wetlands within the urban area are considered as difficult sites in natural cities. If desert land, snow-capped mountains and glaciers, and wetlands are not identified within the urban built-up area, the unused land portion is considered as difficult sites in natural cities. If all three can be identified, they are all considered as difficult sites in natural cities. This depends on the differences in land use types in different urban built-up areas. Through practical testing, urban built-up areas almost do not contain desert land, snow-capped mountains and glaciers, and wetlands. Therefore, the scope of unused land is mainly extracted.

[0048] To prevent omissions in the identification of certain natural urban difficult sites, NDVI was retrieved from 30m Landsat 8 remote sensing imagery data extracted by the U.S. Geological Survey (https: / / earthexplorer.usgs.gov / ). Low-coverage grasslands with NDVI < 0.2 were then added to the list of natural urban difficult sites. The two parts were overlaid to form the extent of the natural urban difficult sites.

[0049]

[0050] In the formula: NIR is the near-infrared band reflectance value, and R is the infrared band reflectance value.

[0051] The third step involves referring to the description of artificially created urban challenging sites to identify plots of land that have been left idle or have a negative impact on the surrounding environment due to restoration or management of land previously used for construction. These plots, after ecological restoration, are primarily used for greening or are identified as green spaces by surface remote sensing. Therefore, remote sensing identification of this type of land can be determined using land use data transfer from two consecutive time periods. Areas currently identified as green spaces by remote sensing that were previously construction land are identified as artificially created urban challenging sites. The land use data used is still from the National Glacier, Permafrost and Desert Scientific Data Center (http: / / www.ncdc.ac.cn). Cultivated land, forest land, grassland, and shrubland are reclassified as green space. The transfer trend and amount of construction land and green space are calculated using a land use transfer matrix. The portion transferred from construction land to green space is identified as artificially created challenging sites.

[0052] Specifically, in a broad sense, urban green space refers to various green spaces within the urban planning area, encompassing all vegetation-covered land, open spaces, and water bodies within that area. However, the identification method for artificially constructed urban challenging sites in this invention is based on the fact that areas currently identified as green space were previously construction land. Further targeted analysis of different types within the green space is not required. Therefore, considering the differences in image interpretation and facilitating the analysis of the overall spatial pattern, the four categories of cultivated land, forest land, grassland, and shrubland in the data classification are reclassified as green space. Construction land and water bodies in the data classification are still identified as construction land and water bodies, respectively. The remaining categories are classified as natural urban challenging sites as previously described. Ultimately, the nine land use types in the original data are reclassified into four major categories: green space, construction land, natural urban challenging sites, and water bodies. The overall transfer trend is then examined and analyzed. The transfer trend and amount of construction land and green space are calculated using a land use transfer matrix. A land use transfer matrix-based method for mining land use change information is discussed. The calculation method is shown in Table 3 below.

[0053] Table 3 Land Use Transition Matrix

[0054]

[0055] In the table: rows represent land use types at time T1, and columns represent land use types at time T2. P ij P represents the percentage of the total land area where land type i was converted to land type j during the period T1-T2; ii P represents the percentage of area where land use type i remained unchanged during the period T1-T2. i+ P represents the percentage of total area for land type i at time T1. +j P represents the percentage of the total area of ​​land use types at time T2. i+ -P iiP represents the percentage decrease in area of ​​land type i during the period T1-T2; +j -P jj This represents the percentage increase in the area of ​​land category j during the period T1-T2.

[0056] Land use data for 2020 and 2021 obtained from the National Data Center for Glacier, Permafrost and Desert Science were imported into ArcGIS 10.7 software. The data was cropped according to the study area using the "Crop" tool. The land use classification was reclassified into four categories: green space, construction land, unused land, and water bodies using the "Reclassify" tool. The raster data was converted into vector data using the "Raster to Vector" tool. The land use transfer overlay analysis was performed on the reclassified data from 2020 and the data from 2021 using the "Intersect" tool. The portion of construction land that was transferred to green space was identified as artificially difficult sites in 2020.

[0057] After completing the previous steps, open the attribute table of the data in ArcGIS 10.7, add area calculation for the category "Construction Land - Green Space", and you can get the area of ​​construction land converted into green space in 2020-2021. This area is the area of ​​difficult artificial urban sites in 2020.

[0058] The fourth step is to merge the identified natural and artificial urban difficult site areas to obtain the complete urban difficult site area.

[0059] Example 2

[0060] like Figure 1 As shown, Figure 1 The left figure shows the land use data of a certain city in 2020 extracted from the urban built-up area of ​​the city. After extraction from the urban built-up area, the land use types within the urban built-up area of ​​the city in 2020 are divided into seven categories: cultivated land, forest land, grassland, shrubland, unused land, construction land and water bodies. It does not include desert land, snow-capped mountains and glaciers, and wetlands. Therefore, it is reclassified into four categories using ArcGIS: green space, unused land (after overlaying low-coverage grassland, it is natural urban difficult site), construction land and water bodies.

[0061] Figure 1 The right side shows the NDVI data obtained by vegetation inversion from Landsat 8 remote sensing images for this range, with values ​​ranging from -1 to 1.

[0062] Unused land and grassland were extracted using the attribute extraction module in ArcGIS. Areas with NDVI values ​​less than 0.2 that overlapped with grassland were extracted. Low-coverage grassland with NDVI < 0.2 was then selected using a mask extraction method and mosaicked with unused land into a new raster to synthesize the overall area of ​​natural-type urban difficult sites in the urban built-up area of ​​a certain city in 2020. Figure 2 As shown.

[0063] To extract information on challenging artificial urban sites in a certain city in 2020, land use classification data for 2020 and 2021 is required. First, the 2020-2021 land use data is reclassified into four categories: green space, unused land, construction land, and water bodies. The portion of construction land that has been converted to green space represents challenging artificial urban sites. Then, the land use transfer matrix method is used in ArcGIS's intersection module to calculate the transfer situation for 2020-2021 based on the reclassified 2020 and 2021 data. Figure 3 As shown.

[0064] Using ArcGIS, the portion of construction land converted to green space was extracted by attribute extraction and merged with the layer of the natural-type urban difficult site map to finally obtain the scope of urban difficult site in a certain city in 2020. Figure 4 As shown.

[0065] Example 3

[0066] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the urban difficult site spatial identification method.

[0067] Example 4

[0068] A non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the aforementioned method for spatial identification of difficult urban sites.

[0069] Example 5

[0070] A computer program product includes a computer program that, when executed by a processor, implements the urban difficult site spatial identification method.

[0071] The above description is merely a preferred embodiment of the present invention, and should not be construed as limiting the scope of the invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method for identifying urban difficult site spaces, characterized in that, Comprising the following steps: According to different cause types and site types, the urban difficult site is divided into two categories: natural type urban difficult site and artificial type urban difficult site; The natural type urban difficult site and the artificial type urban difficult site are integrated with the existing land use classification data, the natural type urban difficult site is identified by the remote sensing image of the existing land use classification data, and the artificial type urban difficult site is identified according to whether it needs to be transferred to construction land; Further, the NDVI is inversed through the remote sensing image data, and the low coverage grassland with NDVI less than 0.2 is supplemented as the newly divided natural type urban difficult site; The identified natural type urban difficult site and artificial type urban difficult site are combined to obtain the complete urban difficult site range; The identification method of the artificial type urban difficult site is to use two continuous time land use transfer data to determine, and to screen the land blocks which are currently identified as green land but were construction land, that is, to identify them as artificial type urban difficult site. Specifically, the land use data used in the identification method of the artificial type urban difficult site comes from the National Glacier Permafrost Desert Scientific Data Center, which classifies land use into cultivated land, forest land, grassland, shrub land, construction land, unused land, water body, desert land, snow mountain glacier land and wetland. The cultivated land, forest land, grassland and shrub land in the data classification are reclassified as green land, the construction land and water body in the data classification are still identified as construction land and water body, and the remaining categories are classified as natural type urban difficult site. Finally, the nine land use types in the original data are reclassified into four categories: green land, construction land, natural type urban difficult site and water body. The transfer trend and amount of construction land to green land are calculated by the land use transfer matrix, and the part transferred from construction land to green land is identified as artificial type difficult site.

2. The urban difficult site space recognition method according to claim 1, characterized in that, The natural type urban difficult site types include saline-alkali land, wasteland, sandy land, bare land and bare rock gravel land, glacier and permanent snow, damaged wetland or water area.

3. The urban difficult site space recognition method according to claim 1, characterized in that, The artificial type urban difficult site types include industrial relocation land, idle land under control, landfill, and building green space.

4. The urban difficult site space recognition method according to claim 1, characterized in that, The existing land use classification data comes from the National Glacier Permafrost Desert Scientific Data Center, which classifies land use into cultivated land, forest land, grassland, shrub land, construction land, unused land, water body, desert land, snow mountain glacier land and wetland. The unused land, desert land, snow mountain glacier land and wetland within the city are considered as natural type urban difficult site.

5. The urban difficult site space recognition method according to claim 4, characterized in that, The identification method of the natural type urban difficult site is to use the existing land use classification data to identify and superimpose the natural type urban difficult site types, further inverse the NDVI through the remote sensing image data, supplement the low coverage grassland with NDVI less than 0.2 as the natural type urban difficult site, and superimpose the two parts to form the natural type urban difficult site range.

6. The urban difficult site space recognition method according to claim 1, characterized in that, The NDVI is inversed according to the Landsat 8 remote sensing image data, and the low coverage grassland data with NDVI less than 0.2 is obtained.

7. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that: The processor executes the program to realize the urban difficult site spatial identification method of any one of claims 1-6.

8. A non-transitory computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by a processor to implement the urban difficult site space identification method according to any one of claims 1-6.

9. A computer program product comprising a computer program, characterized in that, The computer program is executed by a processor to implement the urban difficult site space identification method according to any one of claims 1-6.

Images