CN115661670B

CN115661670B - Method and device for identifying transition region on urban and rural natural landscape gradient

Info

Publication number: CN115661670B
Application number: CN202211598752.3A
Authority: CN
Inventors: 张少尧; 邓伟; 彭立; 刘颖
Original assignee: Sichuan Normal University
Current assignee: Sichuan Normal University
Priority date: 2022-12-14
Filing date: 2022-12-14
Publication date: 2023-04-07
Anticipated expiration: 2042-12-14
Also published as: CN115661670A

Abstract

The invention discloses a method and a device for identifying transition areas on urban and rural natural landscape gradients, wherein the method comprises the following three steps: s1, selecting a characterization factor of human activity intensity, and calculating an HAII index (human activity intensity index) by using a GWPAC analysis model; s2, rearranging the county HAII grids based on a color difference gradual change principle, and extracting transitional curves of the county; and S3, converting the transitional curve into an accumulated dispersion sum curve, calculating the curvature of the accumulated dispersion sum curve, and identifying a transitional area. The method is based on the principal component analysis of geographic weighting, grid sorting, extraction of transitional curves, curve curvature calculation and K-means clustering, constructs a method framework which can be widely applied to identifying the transition areas on the urban and rural natural landscape gradient on different regional scales, analyzes the types of the transition areas, analyzes the transition trend and analyzes the transition area pattern characteristics on the urban and rural natural landscape gradient.

Description

Method and device for identifying transition region on urban and rural natural landscape gradient

Technical Field

The invention belongs to the technical field of landscape gradient recognition and transition area recognition, and particularly relates to a method and a device for recognizing a transition area on an urban and rural natural landscape gradient.

Background

The urban and rural natural gradient can orderly combine the spatial landscapes with different regional functions, show the influence, strength, range and direction of the cross-border adjacency of the urban, rural and natural regional systems, express the dependency relationship and the mutual communication between the natural environment and the social and economic processes, and the dependency relationship and the amplitude and strength of the mutual communication depend on the distribution trend, the characteristic comparison, the combination relationship and the like of the spatial adjacent entities in the gradient sequence. Therefore, the natural gradient of urban and rural areas is not simply to separate the original urban and rural binary space into ternary space, but rather to form a continuous landscape sequence by a series of mutually adjacent and interactive landscapes, wherein the transitional landscapes which link the main landscapes and promote the interaction of landscape functions have been observed and captured by more and more researches. The most studied for transitional landscapes are urban and rural interlaced zones (urban and rural marginal zones) and farming and pasturing (agriculture and forestry) interlaced zones. The urban and rural interlaced zone mainly researches and depicts the regional landscape characteristics of urban and rural activity mixing, spatial superposition and nesting presented in the middle zone from the city to the countryside in the urbanization process. The farming-grazing staggered zone is mainly used for recognizing and analyzing an ecosystem service boundary, a landscape type and an environmental gradient from natural elements, and a fuzzy space boundary has a thick empiric color.

At present, the recognition scheme of urban and rural interlaced zones focuses on recognizing and analyzing transition space and gradient landscape from the perspective of socioeconomic graduality, and the scheme only captures a middle complex interface and transition landscape from urban and rural ends, and because of the urban and rural binary limitations, the recognition and construction of a gradient sequence are incomplete; in the identification scheme of ecological staggered belts such as farming and pasturing staggered belts and the like, the concept is highlighted by the formability of natural environment on landscape gradient and transition space, the decision effect of social and economic activities in local space transformation and reconstruction is not reflected sufficiently, and the ecological staggered belts are especially used for promoting inlaying and stacking of artificial landscape and natural landscape, such as terrace reclamation, abandoned land, seasonal grazing and fallow rotation and the like. In addition, various different transition area identification schemes at the present stage have various defects, such as: 1. the transition area identification technology based on a single open-air observation or interpretation index (land utilization, night light, vegetation index, impervious surface, interest point POI, road network and the like) can only be applied to a research area with a single transition area, for example, only an urban and rural transition area can be identified based on the impervious surface or the interest point POI, but ecological transition zones such as a farming and pasturing transition area and the like cannot be identified, and similarly, only a farming and pasturing transition zone or a farming and forestry transition zone can be identified based on the vegetation index, so that the urban and rural identification effect is poor, and when the research area is large in scale and has various types of transition areas, particularly when the research area faces the transition areas with rich and diverse types on urban and rural natural gradients, the single index has great limitation; 2. although a multi-dimensional index system method based on static linear weighting integrates multiple indexes, the index weight difference in different transition areas cannot be reflected when a large-scale area is faced, and particularly the expression of high time-space heterogeneity in urban and rural transition areas is poor; 3. the density segmentation and mutation detection method based on the fixed threshold has a good identification effect in a single-center circle layer type urban and rural area or farming and animal husbandry linear decreasing area in a plain area, but shows limitation in a multi-center cluster type or flat systematic area in an urban and rural system or a mountain area with complex agriculture, forestry, farming and animal husbandry landscape; 4. although the machine learning method can adapt to the heterogeneity and dynamic expression of a complex region, the initial training and sample definition are required, and the application of the machine learning method in a small-scale region is still feasible, and the machine learning method is applied in a large-scale region and faces high uncertainty.

Meanwhile, the research on the urban and rural natural landscape gradient still stays at the framework level, and the research on large-scale areas is mostly the rough division of space development patterns, rather than from the perspective of landscape geographic space or regional functions. Although few research attempts are made to partition landscape gradient sequences analyzed from urban and rural natural gradient perspectives, transition region spaces on the urban and rural natural gradients are not really identified, and transition region types and trends thereof cannot be analyzed, but the method is very important for enriching social and economic activities and providing elastic spaces for current and future sustainable development and transformation growth.

In conclusion, the identification research of the transition region on the urban and rural natural landscape gradient in the large-scale area range and the research of the type and the trend of the transition region have important significance in the field of sustainable development and landscape evolution research.

Disclosure of Invention

The invention aims to overcome one or more defects in the prior art and provides a method and a device for identifying a transition area on an urban and rural natural landscape gradient.

The purpose of the invention is realized by the following technical scheme:

first aspect

The invention provides a method for identifying a transition area on an urban and rural natural landscape gradient, which comprises the following steps:

selecting a first factor representing the human activity intensity, wherein the first factor comprises cultivated land, construction land, POI (point of interest), road network, night light and population density;

establishing a kilometer grid according to the range of the research area, sampling a first factor by taking each grid unit in the kilometer grid as a unit, and simultaneously assigning a name of a county domain to which each grid unit belongs;

inputting all sampled first factors into a constructed GWPAC analysis model, and modeling by combining a geographical weighted regression algorithm and a principal component analysis algorithm through the GWPAC analysis model;

extracting a plurality of principal components corresponding to each grid unit, the score value of each principal component, the explanation variance ratio of each principal component and the accumulated explanation variance ratio of all principal components by using a GWSCA analysis model, calculating the score weight of each principal component according to the accumulated explanation variance ratio, then respectively carrying out weighted summation on the score value and the score weight of all principal components corresponding to each grid unit, and generating an HAII index corresponding to each grid unit after weighted summation, wherein the HAII index is a human activity intensity index;

converting a kilometer grid of the county area into an original grid layer by taking the county area as a unit, wherein a conversion field comprises HAII indexes, so that HAII grids which are in one-to-one correspondence with all grid units are formed, and reordering HAII index pixel values of all HAII grids according to the size sequence; if the number of lines of the original raster image layer is larger than or equal to the number of columns, placing HAII index pixel values of the reordered HAII raster to a valued area of the original raster image layer line by line, otherwise, placing the HAII index pixel values of the reordered HAII raster to the valued area of the original raster image layer line by line, and generating a new HAII sequencing raster image layer after the placement is finished; if the number of lines of the HAII sequencing grid layer is less than the number of columns, respectively calculating the mean value of HAII index pixel values of each column of HAII grids of the HAII sequencing grid layer, then drawing a transitional curve representing the HAII index change condition of the county according to the column number and the mean value of the HAII index pixel values of each column of HAII grids, otherwise, respectively calculating the mean value of the HAII index pixel values of each row of HAII grids of the HAII sequencing grid layer, and then drawing the transitional curve representing the HAII index change condition of the county according to the row number and the mean value of the HAII index pixel values of each row of HAII grids;

converting the drawn transitional curve of the county area into an accumulated dispersion sum curve by taking the county area as a unit, generating a curvature curve of the accumulated dispersion sum curve based on polynomial simulation and curvature calculation empirical formula, identifying a mutation point, a mutation row and a mutation threshold range of the curvature curve based on an adaptive window method, and determining the HAII grid as a transitional area if the HAII index pixel value of the HAII grid before reordering is within the mutation threshold range, wherein the mutation point is a inflection point with the maximum curvature value.

In a further improvement, the method for identifying the transition area further comprises the following steps:

determining the mutation segment of curvature curve according to the mutation point of curvature curve in county region, and calculating the mutation

And then judging the transition trend of the county transition curve according to the kurtosis and the curvature value of the mutation point.

based on the principle that the color difference gradient represents the transitional geographic space, the corresponding places of the transition region determined by taking a county area as a unit are marked on the grid unit of the kilometer grid of the research area, and the kurtosis and the curvature value of the mutation point of the county area of the transition region are marked on the grid unit marked with the transition region.

selecting a plurality of second factors characterizing the type of transition zone, the plurality of second factors including a topographical index, population density, and GDP;

counting the mean value of various second factors in all grid cells marking the transition region;

and performing cluster analysis on each grid unit marked with the transition region according to the mean value of each second factor, the kurtosis marked by the grid unit and the curvature value of the mutation point, and determining the type of each transition region through the cluster analysis.

In a further improvement, when the name of the county area to which each grid cell belongs is specified, if the area of the grid cell across counties falls into one county area is the largest, the name of the county area is specified as the name of the county area of the grid cell across counties.

In a further improvement, before all the sampled first factors are input into the constructed GWPAC analysis model, a z-score scoring method is used for standardizing each first factor; before extracting a plurality of main components corresponding to each grid unit, the score of each main component, the explanation variance ratio of each main component and the accumulated explanation variance ratio of all the main components by using a GWSCA analysis model, searching an optimal bandwidth in each county, and extracting a plurality of main components corresponding to each grid unit, the score of each main component, the explanation variance ratio of each main component and the accumulated explanation variance ratio of all the main components by using the GWSCA analysis model in the range of the searched optimal bandwidth.

In a further improvement, the plurality of principal components are a first principal component, a second principal component and a third principal component, respectively; when a kilometer grid of a county area is converted into an original grid map layer, the conversion field further comprises score values of all main components, an HAII index forms a first waveband pixel value corresponding to an HAII grid in the original grid map layer, the score value of the first main component forms a second waveband pixel value corresponding to the HAII grid in the original grid map layer, the score value of the second main component forms a third waveband pixel value corresponding to the HAII grid in the original grid map layer, the score value of the third main component forms a fourth waveband pixel value corresponding to the HAII grid in the original grid map layer, and the first waveband pixel value is the HAII index pixel value; before reordering HAII index pixel values of all HAII grids according to the size sequence, performing multi-stage reordering on all wave band pixel values of all grids in an original grid map layer, and forming a sequencing index after sequential comparison, wherein the multi-stage reordering rule is as follows: the first sorting field is first wave band pixel values of all grids in the original grid map layer, the second sorting field is second wave band pixel values of all grids in the original grid map layer, the third sorting field is third wave band pixel values of all grids in the original grid map layer, and the fourth sorting field is fourth wave band pixel values of all grids in the original grid map layer;

and sequentially comparing the positions of the first waveband pixel value before and after the multi-stage reordering to generate a sequencing index.

Further improved, the curvature curve of the cumulative dispersion sum curve is generated based on polynomial simulation and curvature calculation empirical formula, which includes the following sub-steps:

performing polynomial simulation on the accumulated dispersion and the curve to obtain a polynomial function corresponding to the accumulated dispersion and the curve after the polynomial simulation; calculating the first derivative and the second derivative of the polynomial function; calculating curvature curve based on curvature empirical calculation formula

Wherein is present>

Represents the first derivative, -is present>

Representing the second derivative;

the method for identifying the abrupt change point, the abrupt change row and column and the abrupt change threshold range of the curvature curve based on the adaptive window method specifically comprises the following substeps:

constructing a self-adaptive window, comparing whether the curvature value change directions of curvature curves on two sides of the central point of the window are consistent or not by window, and if not, identifying the central point of the window as the inflection point of the curvature curve;

after all inflection points of the curvature curve are identified, defining the inflection point with the maximum curvature value as a catastrophe point, and marking the row number or the column number of the catastrophe point;

identifying a previous row of a row where the mutation point is located to a next row of the row where the mutation point is located as a mutation row, determining an average value of HAII index pixel values corresponding to a row number of the previous row of the row where the mutation point is located as a first threshold value based on a transitional curve, and determining an average value of HAII index pixel values corresponding to a row number of the next row of the row where the mutation point is located as a second threshold value, or identifying a previous column of the column where the mutation point is located to a next column of the column where the mutation point is located as a mutation column, and determining an average value of HAII index pixel values corresponding to a column number of the previous column where the mutation point is located as a first threshold value based on the transitional curve, and determining an average value of HAII index pixel values corresponding to a column number of the next column where the mutation point is located as a second threshold value;

the mutation threshold range is composed of a first threshold and a second threshold, wherein the first threshold is used as a first end point of the mutation threshold range, and the second threshold is used as a second end point of the mutation threshold range.

In a further refinement, the topographic location index

Wherein E represents the elevation on the grid unit marked the transition zone, S represents the slope value of the grid unit marked the transition zone, and/or>

Represents the average elevation, based on the county area to which the grid cell marked the transition area belongs>

Representing an average gradient value of a county area to which the grid cell marking the transition region belongs; peaks labeled according to mean, grid cell of various second factorsAnd when clustering analysis is carried out on each grid unit marked with the transition region by the degree and mutation point curvature value, searching the optimal clustering number from different preset clustering numbers, and then carrying out clustering analysis on the optimal clustering number based on a K-means clustering algorithm.

The first aspect of the invention has the following beneficial effects:

(1) The first aspect of the invention relies on multivariate geographic data (first factor) to generate the human activity intensity index (HAII index), the data compatibility is better, and the adaptive index is wide; meanwhile, a method for combining a geographical weighting regression algorithm with a Principal component analysis algorithm is firstly adopted to measure the human activity intensity index, and the method for combining the geographical weighting regression algorithm with the Principal component analysis algorithm can also be called a Principal component analysis method (GWSCA) based on geographical weighting, so that the divided urban and rural natural landscape gradient is more accurate; in addition, based on the dynamic weight and the dynamic threshold value of the geographic weighting, the method can adapt to areas with high space-time heterogeneity, such as multi-center grouping type, flat systematic areas or mountain areas with complex agriculture, forestry, agriculture and animal husbandry landscapes;

(2) Based on a color difference gradual change principle, namely an action principle of color difference gradual change in transitional geographic space representation, the first aspect of the invention extracts transitional curves in each county area by constructing a grid pixel value reordering technology, so that the identified transitional curves are more accurate, and the identification precision of a transitional area on the urban and rural natural landscape gradient is improved;

(3) Judging the transition trend and the steepness degree of a transition curve through the kurtosis and the curvature value of the catastrophe point, wherein the larger the kurtosis and the curvature value of the catastrophe point is, the stronger the HAII index transition in the county area is, and on the contrary, the smaller the curvature and the curvature value of the catastrophe point is, the gentler and smoother the HAII index transition in the county area is, so that on the basis of identifying the transition region on the natural landscape gradient of the town and country, the judgment on the transition trend is also realized, and the sustainable development and landscape evolution research field can be further supported;

(4) The type of the transition region is identified through cluster analysis, the classification of the transition region is realized, and meanwhile, the accuracy of the classification of the transition region is further improved based on the use of multiple second factors;

(5) Compared with the prior art that only a single urban and rural transition area or farming and pasturing transition area can be identified, the first aspect of the invention realizes successful identification of different types of transition areas on continuous landscape gradients in a large-scale area;

(6) The county area is used as an analysis unit, different types of transition areas inside the county area can be identified, the limitation that the identified transition areas are mostly concentrated near the geographic boundary in the prior art is broken through, and therefore the transition areas identified by the method are more comprehensive and fit with the actual natural environment and social economic development process.

Second aspect of the invention

The second aspect of the invention provides a device for identifying a transition area on an urban and rural natural landscape gradient, which comprises a memory and a processor, wherein the memory stores the method for identifying the transition area on the urban and rural natural landscape gradient provided by the first aspect of the invention, and the processor is used for calling the method for identifying the transition area on the urban and rural natural landscape gradient stored in the memory to identify the transition area.

The second aspect of the present invention brings about the same advantageous effects as the first aspect, and will not be described in detail herein.

Drawings

FIG. 1 is a flow chart of a method for identifying transition areas on the gradient of natural landscapes in urban and rural areas;

FIG. 2 is a schematic diagram of a spatial distribution of urban and rural natural landscape gradients in the southwest region;

FIG. 3 is a schematic diagram of a transition region on the gradient of natural urban and rural landscapes in the southwest region;

fig. 4 is a schematic diagram of transition region types and trend patterns on the urban and rural natural landscape gradient in the southwest region.

Detailed Description

The technical solutions of the present invention will be described clearly and completely with reference to the following embodiments, and it should be understood that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be obtained by a person skilled in the art without inventive effort based on the embodiments of the present invention, are within the scope of the present invention.

Example one

Referring to fig. 1 to 4, the embodiment provides a method for identifying a transition area on an urban and rural natural landscape gradient, which is used for identifying the transition area on the urban and rural natural landscape gradient.

Specifically, the method for identifying the transition area on the urban and rural natural landscape gradient comprises the following steps:

s1, selecting a characterization factor of human activity intensity, and calculating an HAII index (human activity intensity index) by using a GWPAC analysis model.

Specifically, S1 includes the following substeps:

s11, selecting a first factor representing the human activity intensity. Wherein the first factor comprises: arable land, construction land, POI, road network, night light and population density.

S12, establishing a kilometer grid according to the range of the research area, sampling the first factor by taking each grid unit in the kilometer grid as a unit, and simultaneously assigning a name of a county area to which each grid unit belongs. Wherein, the kilometer grid refers to a vector sampling grid with the resolution of 1 × 1 km. For vector data, a planar kilometer grid is used to count the number of POIs and the length of roads in the road network within each grid cell, where the multiline roads have been simplified to centerline roads. Through sampling of kilometer grids, the six types of first factor data are converted into a unified multivariate geographic data sampling table, and county domain names (county-level administrative unit names) are superposed on each grid unit. When the name of the county area to which each grid cell belongs is designated, if the area of the grid cell across the counties is the largest, the name of the county area is designated as the name of the county area of the grid cell across the counties. In the present embodiment, the ArcGIS software is preferably used for sampling and extracting the first factor.

And S13, inputting all the sampled first factors into the constructed GWPAC analysis model, and modeling the GWPAC analysis model by combining a geographical weighted regression algorithm and a principal component analysis algorithm. Before all the sampled first factors are input into the constructed GWPAC analysis model, each first factor is normalized by using a z-score scoring method. In this embodiment, the GWPCA analysis model is a GWmodel data packet based on the R language.

S14, extracting a plurality of principal components corresponding to each grid unit, the score value of each principal component, the explanation variance ratio of each principal component and the accumulated explanation variance ratio of all principal components by using a GWPAC analysis model, calculating the score weight of each principal component according to the accumulated explanation variance ratio, then respectively carrying out weighted summation on the score value and the score weight of all principal components corresponding to each grid unit, and generating HAII indexes corresponding to each grid unit after weighted summation, wherein the HAII indexes are human activity intensity indexes. And (4) dividing the urban and rural natural landscape gradient by calculating the human activity intensity index.

Optionally, in consideration of irregularity of county domain shapes, before extracting a plurality of principal components, score values of the principal components, explanation variance ratios of the principal components, and accumulated explanation variance ratios of all principal components corresponding to each grid unit by using a GWPCA analysis model, an optimal bandwidth is searched in each county domain, when the optimal bandwidth is searched, an adaptive bandwidth searching method is used, a weight kernel function is a gaussian function, an optimal bandwidth is selected for each county domain by using a cross validation method, and then a plurality of principal components, score values of the principal components, explanation variance ratios of the principal components, and accumulated explanation variance ratios of all principal components corresponding to each grid unit are extracted in the searched optimal bandwidth range.

S2, rearranging the county HAII grid based on the color difference gradual change principle, and extracting a transitional curve of the county.

Specifically, S2 includes the following substeps:

s21, with a county area as a unit, converting a kilometer grid of the county area into an original grid layer based on a principle that a transitional geographic space is represented by chromatic aberration gradual change, wherein a conversion field comprises HAII indexes, so that HAII grids corresponding to all grid units one by one are formed, and then reordering HAII index pixel values of all HAII grids according to a size sequence; if the number of lines of the original raster image layer is larger than or equal to the number of columns, placing HAII index pixel values of the reordered HAII raster to a value area of the original raster image layer line by line, otherwise, placing the HAII index pixel values of the reordered HAII raster to the value area of the original raster image layer line by line, and generating a new HAII sequencing raster image layer after the placement is finished; if the number of lines of the HAII sequencing grid layer is smaller than the number of columns, respectively calculating the mean value of HAII index pixel values of each column of HAII grids of the HAII sequencing grid layer, then drawing a transitional curve representing the HAII index change condition of the county area according to the column number and the mean value of the HAII index pixel values of each column of HAII grids, otherwise, respectively calculating the mean value of the HAII index pixel values of each row of HAII grids of the HAII sequencing grid layer, and then drawing the transitional curve representing the HAII index change condition of the county area according to the row number and the mean value of the HAII index pixel values of each row of HAII grids. The drawing process of the transitional curve specifically comprises the following steps: drawing a curve by taking the row number as an abscissa and the mean value of HAII index pixel values of all HAII grids corresponding to each row as an ordinate, wherein the curve is a transitional curve representing the HAII index change condition of the county area; or; and drawing a curve by taking the column number as an abscissa and taking the mean value of the HAII index pixel values of all HAII grids corresponding to each column as an ordinate, wherein the curve is a transitional curve representing the HAII index change condition of the county area.

And placing the reordered HAII raster HAII index pixel values only in the valued area of the original raster image layer, thereby ensuring that the space range of the reordered new HAII ordered raster image layer is consistent with that of the original raster image layer, and generating the HAII ordered raster image layer after reordering as an HAII single-waveband raster image layer. In this embodiment, when the HAII index pixel values of the HAII grids are reordered according to the size sequence, it is preferable to reorder the HAII index pixel values of the HAII grids in the order from large to small.

Optionally, when a county-area kilometer grid is converted into an original grid map layer, the conversion field further includes score values of the respective principal components, the HAII index constitutes a first band pixel value corresponding to the HAII grid in the original grid map layer, the score value of the first principal component constitutes a second band pixel value corresponding to the HAII grid in the original grid map layer, the score value of the second principal component constitutes a third band pixel value corresponding to the HAII grid in the original grid map layer, the score value of the third principal component constitutes a fourth band pixel value corresponding to the HAII grid in the original grid map layer, and the first band pixel value is the HAII index pixel value. Before reordering HAII index pixel values of all HAII grids in a descending order, performing multi-level reordering on all wave band pixel values of all grids in an original grid map layer, wherein the multi-level reordering rule is as follows: the first ordering field is first band pixel values of all HAII grids in the original grid map layer, the second ordering field is second band pixel values of all HAII grids in the original grid map layer, the third ordering field is third band pixel values of all HAII grids in the original grid map layer, and the fourth ordering field is fourth band pixel values of all HAII grids in the original grid map layer; and generating an ordering index after sequentially comparing the positions of the first waveband pixel value before and after the multi-stage reordering.

And S3, converting the transitional curve into an accumulated dispersion sum curve, calculating the curvature of the accumulated dispersion sum curve, and identifying a transitional area.

Specifically, S3 includes the following substeps:

s31, with a county area as a unit, converting the drawn transitional curve of the county area into an accumulated dispersion sum curve (CUSUM curve), generating a curvature curve of the accumulated dispersion sum curve based on polynomial simulation and a curvature calculation empirical formula, and identifying a mutation point, a mutation row and a mutation threshold range of the curvature curve based on an adaptive window method.

Alternatively, the conversion formula for converting the transitional curve into the cumulative dispersion sum curve is:

，/>

is a cumulative sum of the ith row or column>

Is a cumulative sum of a row preceding the ith row or a column preceding the ith column>

Is the mean of the HAII index pixel values in row i or column i->

Is the mean of all ordinate data of the entire transition curve.

Optionally, generating a curvature curve of the cumulative dispersion sum curve based on a polynomial simulation and a curvature calculation empirical formula specifically includes the following sub-steps: performing polynomial simulation on the accumulated dispersion and the curve to obtain a polynomial function corresponding to the accumulated dispersion and the curve after the polynomial simulation; calculating the first derivative and the second derivative of the polynomial function; calculating curvature curve based on curvature empirical calculation formula

In which>

Represents a first derivative, <' > based on>

The second derivative is indicated. Since the polynomial function can better simulate the trend of curve variation, the fourth-order polynomial function is used to simulate the cumulative dispersion and curve in the present embodiment, and the accuracy (R) is simulated ² ) Above 0.9.

Optionally, identifying a break point, a break line and a break threshold range of the curvature curve based on an adaptive window method specifically includes the following sub-steps:

constructing a self-adaptive window, comparing whether the curvature value change directions of curvature curves at two sides of the center point of the window are consistent or not by window, and if not, identifying the center point of the window as the inflection point of the curvature curve;

and after all inflection points of the curvature curve are identified, defining the inflection point with the maximum curvature value as a catastrophe point, and marking the row number or the column number of the catastrophe point. When all inflection points of the curvature curve are identified according to the curvature meaning, the starting position and the ending position of the curvature curve are excluded;

identifying a previous row of a row where the mutation point is located to a next row of the row where the mutation point is located as a mutation row, determining the average value of HAII index pixel values corresponding to the row number of the previous row of the row where the mutation point is located (the row number of the previous row of the row where the mutation point is located is an abscissa, and a corresponding ordinate value on a transitional curve) as a first threshold value based on a transitional curve, and determining the average value of HAII index pixel values corresponding to the row number of the next row of the row where the mutation point is located (the row number of the next row of the row where the mutation point is located is an abscissa, and a corresponding ordinate value on a transitional curve) as a second threshold value, or identifying a previous column of the column where the mutation point is located to a next column where the mutation point is located as a mutation column, and determining the average value of HAII index pixel values corresponding to the column number of the previous column of the mutation point as a first threshold value based on a transitional curve, and determining the average value of HAII index pixel values corresponding to the column number of the next column of the mutation point as a second threshold value;

the mutation threshold range is composed of a first threshold and a second threshold, wherein the first threshold is used as a first end point of the mutation threshold range, and the second threshold is used as a second end point of the mutation threshold range. For example: and if the mutation point is on the 15 th row, identifying the 14 th to 16 th rows as mutation rows, wherein in the transitional curve of the county area, the mean value of the HAII index pixel values corresponding to the 14 th row is the first threshold, and the mean value of the HAII index pixel values corresponding to the 16 th row is the second threshold.

And S32, if the HAII index pixel value of the HAII grid before reordering is within the mutation threshold range, determining the HAII grid as a transition region. The method comprises the following specific steps: this abrupt threshold range is returned to the original raster image layer before reordering per county area, and the HAII grid is identified as a transition region if the HAII index pixel value of the HAII grid falls within the abrupt threshold range.

As shown in fig. 2, the method for identifying the transition area implemented in this embodiment is applied to the city of the three provinces in southwest province (the lower city)The disclosure refers to the three provinces-one city in southwest as the southwest region) to identify transition regions on the urban and rural natural landscape gradient in the southwest region. The three provinces in the southwest are respectively: sichuan province, yunnan province and Guizhou province, and the city is Chongqing city. The plurality of geographic data sampling tables comprises 438 sub-county sampling tables. In the whole southwest area, when the GWPCA analysis model is used for carrying out principal component analysis based on geographical weighting on the six first factors, after a plurality of tests, when three principal components are extracted, the cumulative explanation variance ratio reaches more than 90%, so that the number of the extracted principal components in each county area is determined to be three, and the three principal components are respectively a first principal component, a second principal component and a third principal component. The weighted sum formula in calculating the HAII index is as follows:

HAII denotes the HAII index, score1 denotes the score value of the first principal component, ` H `>

Represents the ratio of the interpreted variances of the first principal component, score2 is the score value of the second principal component, and->

Represents the interpretation variance ratio of the second principal component, score3 is the score value of the third principal component, and->

Represents the interpretation variance ratio of the third principal component, and local _ var represents the cumulative interpretation variance ratio.

Analysis of the HAII index obtained gave: the average value of the HAII index in the whole southwest area is 0.79, the standard deviation is 1.46, and the whole human activity intensity is low; the overall distribution pattern of the HAII index presents a distribution pattern with a high east, a low west and a high north, a low south; the Sichuan basin is a region with a higher HAII index in the southwest region, and the HAII indexes of the Tibet plateau in the west, the cross-over mountain region and the Yunobi plateau in the south are lower; the four major cities in the southwest region are the areas with the highest HAII index, which is substantially greater than 20; the HAII index of the vast rural area is lower than 10; the HAII index of most areas of the Qinghai-Tibet plateau and the transverse mountainous area is lower than 0.5, and the intensity of human activities is extremely low.

According to the regional function of the southwest region, and in combination with a density segmentation and natural discontinuity method, the HAII index is divided into seven grades of three types, wherein the three types respectively correspond to a natural region, a rural region and an urban region.

The natural areas are widely distributed in western and south mountain areas, and the area reaches 62.43 km ² More than half of the area of the southwest region (54.13%), the HAII index of the middle and western intersecting mountain regions in the natural region is the lowest, the HAII index of most regions is lower than 0.5, the trace and the strength of the human motion are extremely low, and the natural reserve land N1 belongs to; meanwhile, in the south cloud plateau and the north basin periphery mountain land, the HAII index is lower than 1.5, the humanistic activity footprint and the strength are higher than those of a natural reserve land N1, and partial regions are distributed with remote rural settlement and belong to a region N2 transiting from the natural reserve land to the rural regions.

The rural areas can be divided into two types according to the HAII index, namely an R1 area and an R2 area which account for 2.20 percent of the southwest area; wherein, the R1 area with HAII index less than 5 is the main distribution area (area ratio is 1.42%) of the country, the main distribution area is intensively distributed in Sichuan basin, the concentrated distribution areas on the cloud plateau are scattered, and the R1 area of the cross mountain area is mainly distributed along the valley and the road; the R1 area is a main distribution area of rural settlement and cultivated land and bears the regional functions of rural residence and agricultural production; the R2 areas are less in occupation and mainly distributed in the peripheral areas of various large urban areas, such as Chengdu, chongqing, guiyang and Kunming, are distributed in the continuous R2 areas, belong to the transition area from villages to cities, and are the areas with the most obvious urban and rural transformation change in recent years.

The urban area percentage of the southwest area is 0.47%, and the distribution pattern is more south-east and less north-west. Classifying the urban areas into three types according to the HAII index, wherein the three types are a U1 area, a U2 area and a U3 area; the U1 area is distributed most widely, and is a main component of each large, medium and small city and a suburban component of four main cities throughout the southwest area; the U2 area accounts for only 0.08%, is the central urban area of each local city, and is urbanizedRegions of relative maturity develop; the area of the U3 area is only 214 km ² The percentage is 0.02%, and the method is mainly distributed in the central urban areas of Chengdu, chongqing, guiyang and Kunming, is the area with the highest maturity of urbanization development, and is the economic and cultural center in the southwest area.

And according to the mutation threshold range extracted from each county curvature curve, defining the grid unit with the HAII index within the mutation threshold range in the grid unit of the southwest region as a transition region, thereby extracting the transition region of the whole southwest region.

Example two

The difference between the second embodiment and the first embodiment is that: the following steps are also included after step S3:

and S4, calculating the kurtosis in the transition region, and judging the transition trend of the transitional curve in the county according to the kurtosis and the catastrophe point curvature value. The method specifically comprises the following steps: determining a mutation segment of the curvature curve according to a mutation point of the county curvature curve by taking the county as a unit, calculating the kurtosis of the curvature of the mutation segment, and judging the transition trend of the county transitional curve according to the kurtosis and the curvature value of the mutation point.

Wherein, the catastrophe point A is searched according to the inflection point identified by the curvature curve of the county area _n And the mutation point A _n Adjacent previous inflection point A _n-1 With the next inflection point A _n+1 Marking inflection points A respectively _n-1 And an inflection point A _n+1 The corresponding line number or column number, the inflection point A _n-1 And inflection point A _n+1 The interval of curvature values corresponding to the row number or the column number of the group is used as the abrupt change [ C _n-1 ~C _n+1 ]Wherein, C _n-1 Is an inflection point A _n-1 Curvature value of (C) _n+1 Is an inflection point A _n+1 The curvature value of (a). If the curvature curve of the county area has only one inflection point, the whole curvature value sequence is regarded as a mutation segment, if the mutation point A _n Is the first inflection point of the curvature curve, the first data point of the curvature value sequence is considered as the adjacent inflection point by default, if the discontinuity point a _n Being the last inflection point of the curvature curve, the last data point of the curvature value sequence is considered as an adjacent inflection point by default. After the mutation segment is determined, useValue of curvature [ C ] in abrupt transition _n-1 ~C _n+1 ]The kurtosis statistic of (a) characterizes the sharpness and steepness of the curvature distribution at the mutation point. Measuring the transition trend and the steepness degree of a transitional curve in a county by using the kurtosis and the curvature value of the mutation point in the mutation segment together, wherein the transition trend of the transitional curve is positively correlated with the kurtosis and the curvature value of the mutation point in the mutation segment, namely: the bigger the kurtosis and mutation point curvature values are, the more violent the transition of the transitional curve in county area is shown; conversely, a smaller kurtosis versus inflection point curvature value indicates a more gradual and smooth transition of the transitional curve within the county.

And S5, identifying different transition region types by using a K-means clustering algorithm according to the kurtosis, the catastrophe point curvature value and a second factor for representing the transition region types. The method specifically comprises the following steps: based on the principle that the color difference gradient represents transitional geographic space, marking the corresponding places of the transition region determined by taking a county area as a unit on grid units of a kilometer grid of a research area, and marking the kurtosis and the curvature value of a mutation point of the county area to which the transition region belongs on the grid units marked with the transition region; selecting a plurality of second factors for representing the types of the transition areas, wherein the plurality of second factors are terrain location indexes, population densities and GDPs; counting the average value of various second factors in all grid cells marking the transition area; and performing cluster analysis on each grid unit marked with the transition region according to the mean value of each second factor, the kurtosis marked by the grid unit and the curvature value of the mutation point, and determining the type of each transition region through the cluster analysis.

Optionally, a terrain position index

E represents the elevation on the grid unit marked the transition zone, S represents the slope value of the grid unit marked the transition zone, and->

RepresentMarking the average gradient value of the county domain to which the grid unit of the transition region belongs;

optionally, when clustering analysis is performed on each grid unit marked with the transition region according to the mean value of each second factor, the kurtosis marked by the grid unit and the curvature value of the mutation point, the optimal clustering number is searched from different preset clustering numbers, and then the clustering analysis of the optimal clustering number is performed based on a K-means clustering algorithm. The preset different cluster numbers are respectively as follows: 2. 3, 4, 5, 6, 7, 8, 9 and 10, and when the optimal cluster quantity is searched, selecting the optimal cluster quantity according to the cluster evaluation scores corresponding to various cluster quantities.

After the types of the transition areas are identified, the trends and the types of the transition areas on the urban and rural natural landscape gradients can be statistically analyzed, then the regional characteristics, the spatial distribution patterns and the modes of the transition areas of different types are summarized, the kurtosis and the inflection point curvature values of the transition areas of different types are compared, more different transition trends are revealed, and meanwhile, the method can be used for providing planning inspiration for geographic planning, urban planning, ecological service management and the like.

As shown in fig. 3, the classification method for the transition region type implemented in this embodiment is applied to the southwest region, and the following analysis results are obtained:

where each point in fig. 3 represents a transition region of 1 square kilometer.

The area of the transition region in the southwest area is 61880 km ² And occupies about 5.37% of the total area. The transition area is distributed uniformly in the Sichuan basin and the Yunobao plateau, but presents a gathering distribution situation in the cross mountain area, and is mainly distributed along the valley and the road. The transition areas in the southwest area can be divided into four types, and the four types of transition areas are respectively combined with the spatial distribution of each type of transition area and the urban and rural natural gradient: suburban transition areas (M-STZ), urban and rural transition areas (U-RTZ), humanitarian dominant mountainous area transition areas (MTZ-H) and natural dominant mountainous area transition areas (MTZ-N);

the suburban transition zone (M-STZ) is a transition zone between the major urban metropolitan area and suburban area, and is the forefront of urbanization expansion. The M-STZ area is minimum, and is only 406 km ² About 0.66% of the entire transition zone area, mainly distributedIn the periphery of major cities such as Chengdu, chongqing, guiyang and Kunming, the M-STZ area of the Chengdu city is the largest and reaches 195 km ² And accounts for 48.03 percent of the area of the M-STZ. The M-STZ is mainly distributed in a U1 area on the urban and rural natural gradient, accounts for 95.81 percent of the whole M-STZ, and the U2 area accounts for a very small amount, which shows that the M-STZ mainly reflects HAII index transition in a urban continuous area;

the urban and rural transition zone (U-RTZ) is a transition zone between urban and rural areas and is the most direct and significant area for urban and rural transformation. The U-RTZ area is 30544 km ² Accounting for 49.36% of the transition zone, is the largest one of the four types of transition zones. The U-RTZ is mainly distributed in the Sichuan basin, the east of Guizhou province and the south of Yunnan province, and is scattered in the north of Yunnan and the south of Sichuan province, and the whole distribution has a distribution pattern with less east-docetaxel. The U-RTZ is mainly distributed in an R1 area and an R2 area on the urban and rural natural gradient and respectively accounts for 70.34 percent and 27.00 percent of the transition area, and simultaneously accounts for 0.17 percent, 1.20 percent and 1.30 percent in the U1 area, the U2 area and the N2 area respectively, so that the U-RTZ is the only transition area which spans the complete urban and rural natural gradient.

The humanitarian mountain transition zone (MTZ-H) is a transition zone between rural areas and nature in the mountain area, and represents the transition from agricultural planting to natural protection. The MTZ-H area is second to the U-RTZ area and reaches 23770 km ² And accounts for 38.41% of the transition region. MTZ-H is mainly distributed in the cloud plateau, the southwest Sichuan is also provided with a centralized distribution area, and the mountainous areas around the Sichuan basin are also distributed, for example, the transition area from the Sichuan basin to the cross mountainous area is also provided with long and narrow distribution. In urban and rural natural gradient, MTZ-H is mainly distributed in N2 area, and occupies 74.79% of the transition region, and in R2 area, the distribution is 16.82%, which shows typical rural-natural transition. The transition region is mainly distributed in the mountainous region with the altitude of over 1000 m, so that the transition property between the agricultural activities in the mountainous region and the ecological environment protection in the mountainous region can be reflected better.

The natural dominant mountain transition zone (MTZ-N) is a transition zone between villages and nature, but more shows the interior of nature. MTZ-N area 7160 km ² 11.57% of the transition zone, mainly distributed in the transverse mountainous areas of the west of Sichuan and the north of Yunnan, and scattered in the south of Sichuan and the north of southeast of YunnanAnd (4) distribution. MTZ-N is mainly distributed in an N1 area on a natural gradient of cities and towns, accounts for 87.56 percent of the transition area, and accounts for 8.81 percent and 3.63 percent of the transition area in an N2 area and an R1 area respectively. The terrain of the MTZ-N distribution area is mostly a plateau high mountain area, the human activity intensity is low, and the transition between a remote country and a natural reserve land is mainly reflected.

As shown in fig. 4, the determination of the transition trend implemented in this embodiment is applied to the southwest region, and the following determination results are obtained:

the HAII index gradually decreases from a super metropolitan area to a natural reserve area, covers nine non-transition areas with different regional functions and transition areas with different transition trends, and forms a complete and diverse transition pattern on a natural gradient of the urban and rural areas in the southwest area, which is composed of a super metropolitan area, a suburban transition area, a vast middle-small city, a rural transition area, a rural agricultural area, a humanitarian mountain transition area, a mountain rural area, a natural dominant mountain transition area and a natural reserve area.

The transition band spectrum of the southwest region is firstly a super metropolitan area, which is a region with the highest HAII index in the southwest region; the second place is a suburb transition zone (M-STZ) which belongs to a region type of transition from a high-density central built-up zone to a low-density suburb and a satellite city, and is a first-class transition zone, the transition trend is mild, the whole transition curve is linear, the curvature curve is short and mild, and the kurtosis is small; the third place is a medium and small city group which is mainly distributed in central urban areas and county-area towns of various local cities in the southwest region; the fourth position is an urban and rural transition zone (U-RTZ), the transition trend of the urban and rural transition zone is obvious, the transition curve is in a linear trend in the first half part and then is in a gentle area, the curvature curve shows a trend of partial normal distribution, and the curvature value and the kurtosis of the catastrophe point are higher than those of the suburban transition zone; the fifth is a rural agricultural area which is mainly distributed in hilly areas of Sichuan basin and flat dam areas of Yungui plateau and mainly belongs to R1 area in HAII index; the sixth is a humanity leading mountain area transition area (MTZ-H), which is a third transition area, the MTZ-H transition trend is rapid, the integral presents a logarithmic curve descending trend, the curvature distribution is more concentrated, the curve distribution is sharper, and the kurtosis is larger; the seventh position is a rural mountainous area which is an area with higher human activities in the mountainous area, the area is a main poor village distribution area in the southwest area and is also a minority rural concentration area, the terrain is rugged, and the traffic accessibility is poor; the eighth is a naturally dominant mountain area transition area (MTZ-N) which is also a fourth-class transition area, the MTZ-N transition trend is the most severe, a transition curve is at an extremely low level and approaches to 0 after linearly descending at the beginning, the curvature curve is the sharpest, and the curvature value and the kurtosis of a catastrophe point are the largest in the fourth-class transition area; the ninth position is a natural reserve area, belongs to a limited/forbidden human activity area, is mainly distributed in a cross mountain area and other high mountain areas, is mostly a national or provincial protection area, has an altitude of more than 4000 meters, and is rarely used by human.

The five non-transition areas and the four transition areas jointly form an urban and rural natural transition band spectrum with various functions, various types and integrity in the southwest area, and the gradient change and combination relation of the social-economic and natural ecological component ratio in the social-ecological coupling system is shown. In the first type of transition zone M-STZ, socioeconomic components occupy an absolute dominant position in a social-ecological coupling system, and few natural ecological components mainly provide a green space environment and a rest space for a city; the social and economic components in the second type of transition region U-RTZ are slightly higher than natural ecological components, wherein the social and economic components are rapid urban and rural transformation landscapes, such as small and medium towns and concentrated continuous ploughing areas, and the natural ecological components are natural reserved landscapes in rural transformation and reconstruction; natural ecological components in a third transition region MTZ-H begin to exceed socioeconomic components, the natural ecological components mainly comprise natural or artificial reserved landscapes in mountainous regions such as forests, grasslands, shrubs and the like, and the socioeconomic components mainly comprise landscapes such as rural settlement and mountainous region cultivated land and the like; natural ecological components in the MTZ-N in the fourth column of transition area occupy an absolute position, social and economic components occupy a small proportion, and the natural elements such as grasslands, forests, rivers and the like mainly form artificial landscapes such as remote villages, pastures and mountain roads and serve as landscape substrates.

EXAMPLE III

The embodiment provides a device for recognizing transition areas on urban and rural natural landscape gradients, which comprises a memory and a processor, wherein the method for recognizing transition areas on urban and rural natural landscape gradients provided by the first embodiment or the second embodiment is stored in the memory, and the processor is used for calling the method for recognizing transition areas on urban and rural natural landscape gradients stored in the memory to recognize the transition areas.

The foregoing is illustrative of the preferred embodiments of this invention, and it is to be understood that the invention is not limited to the precise form disclosed herein and that various other combinations, modifications, and environments may be resorted to, falling within the scope of the concept as disclosed herein, either as described above or as apparent to those skilled in the relevant art. And that modifications and variations may be effected by those skilled in the art without departing from the spirit and scope of the invention as defined by the appended claims.

Claims

1. A method for identifying transition areas on urban and rural natural landscape gradients is characterized by comprising the following steps:

establishing a kilometer grid according to the range of a research area, sampling a first factor by taking each grid unit in the kilometer grid as a unit, and simultaneously designating a name of a county domain to which each grid unit belongs;

inputting all sampled first factors into a constructed GWSCA analysis model, wherein the GWSCA analysis model is combined with a geographical weighted regression algorithm and a principal component analysis algorithm for modeling;

extracting a plurality of principal components corresponding to each grid unit, the score value of each principal component, the explanation variance ratio of each principal component and the accumulated explanation variance ratio of all principal components by using a GWPAC analysis model, calculating the score weight of each principal component according to the accumulated explanation variance ratio, then respectively carrying out weighted summation on the score value and the score weight of all principal components corresponding to each grid unit, and generating an HAII index corresponding to each grid unit after weighted summation, wherein the HAII index is a human activity intensity index;

converting a kilometer grid of the county area into an original grid layer by taking the county area as a unit, wherein a conversion field comprises HAII indexes, so that HAII grids which are in one-to-one correspondence with all grid units are formed, and reordering HAII index pixel values of all HAII grids according to the size sequence; if the number of lines of the original raster image layer is larger than or equal to the number of columns, placing HAII index pixel values of the reordered HAII raster to a value area of the original raster image layer line by line, otherwise, placing the HAII index pixel values of the reordered HAII raster to the value area of the original raster image layer line by line, and generating a new HAII sequencing raster image layer after the placement is finished; if the number of lines of the HAII sequencing raster image layer is less than the number of columns, respectively calculating the mean value of HAII index pixel values of each column of HAII grids of the HAII sequencing raster image layer, then drawing a transitional curve representing the HAII index change condition of the county area according to the column number and the mean value of the HAII index pixel values of each column of HAII grids, otherwise, respectively calculating the mean value of the HAII index pixel values of each row of HAII grids of the HAII sequencing raster image layer, and then drawing a transitional curve representing the HAII index change condition of the county area according to the row number and the mean value of the HAII index pixel values of each row of HAII grids;

2. The method for identifying the transition area on the urban and rural natural landscape gradient according to claim 1, wherein the method for identifying the transition area further comprises the following steps:

determining an abrupt change section of the curvature curve according to an abrupt change point of the county curvature curve by taking the county as a unit, calculating the kurtosis of the curvature value of the abrupt change section, and judging the transition trend of the county transitional curve according to the kurtosis and the curvature value of the abrupt change point.

3. The method for identifying the transition area on the urban and rural natural landscape gradient according to claim 2, wherein the method for identifying the transition area further comprises the following steps:

based on the principle that the color difference gradient represents the transitional geographic space, the corresponding places of the transition region determined by taking the county area as a unit are marked on grid units of a kilometer grid of the research region, and the kurtosis and the curvature value of the mutation point of the county area of the transition region are marked on the grid units marked with the transition region.

4. The method for identifying transition areas on urban and rural natural landscape gradients as claimed in claim 3, wherein the method for identifying transition areas further comprises the following steps:

and performing cluster analysis on each grid unit marked with the transition region according to the mean value of each second factor, the kurtosis marked by the grid unit and the curvature value of the catastrophe point, and determining the type of each transition region through the cluster analysis.

5. The method as claimed in claim 1, wherein when the name of the county area is assigned to each grid cell, if the area of the grid cell across the county area is the largest, the name of the county area is assigned to the county area of the grid cell across the county area.

6. The method for identifying transition areas on urban and rural natural landscape gradients as claimed in claim 1,

before all sampled first factors are input into the constructed GWPAC analysis model, normalizing each first factor by using a z-score scoring method;

before extracting a plurality of main components corresponding to each grid unit, the score of each main component, the explanation variance ratio of each main component and the accumulated explanation variance ratio of all the main components by using a GWSCA analysis model, searching an optimal bandwidth in each county, and extracting a plurality of main components corresponding to each grid unit, the score of each main component, the explanation variance ratio of each main component and the accumulated explanation variance ratio of all the main components by using the GWSCA analysis model in the range of the searched optimal bandwidth.

7. The method for identifying transition areas on urban and rural natural landscape gradients as claimed in claim 1,

the plurality of principal components are a first principal component, a second principal component and a third principal component, respectively;

when a kilometer grid of a county area is converted into an original grid map layer, the conversion field further comprises score values of all main components, an HAII index forms a first waveband pixel value corresponding to an HAII grid in the original grid map layer, the score value of the first main component forms a second waveband pixel value corresponding to the HAII grid in the original grid map layer, the score value of the second main component forms a third waveband pixel value corresponding to the HAII grid in the original grid map layer, the score value of the third main component forms a fourth waveband pixel value corresponding to the HAII grid in the original grid map layer, and the first waveband pixel value is the HAII index pixel value;

before reordering HAII index pixel values of all HAII grids according to the size sequence, performing multi-stage reordering on all wave band pixel values of all grids in an original grid graph layer, wherein the multi-stage reordering rule is as follows: the first sorting field is first wave band pixel values of all grids in the original grid map layer, the second sorting field is second wave band pixel values of all grids in the original grid map layer, the third sorting field is third wave band pixel values of all grids in the original grid map layer, and the fourth sorting field is fourth wave band pixel values of all grids in the original grid map layer;

8. The method for identifying transition areas on urban and rural natural landscape gradients as claimed in claim 1,

the curvature curve of the cumulative dispersion sum curve is generated based on polynomial simulation and a curvature calculation empirical formula, and the method specifically comprises the following substeps:

performing polynomial simulation on the accumulated dispersion sum curve to obtain a polynomial function corresponding to the accumulated dispersion sum curve; calculating the first derivative and the second derivative of the polynomial function; calculating curvature curve based on curvature empirical calculation formula

In which>

Represents the first derivative, -is present>

Representing the second derivative;

9. The method for identifying transition areas on urban and rural natural landscape gradients as claimed in claim 4,

the topographic and topographic indexes

Wherein E represents the elevation on the grid unit marking the transition zone, S represents the slope value of the grid unit marking the transition zone, and->

Representing an average gradient value of a county area to which the grid cell marking the transition region belongs;

and when clustering analysis is carried out on each grid unit marked with the transition region according to the mean value of each second factor, the kurtosis marked by the grid unit and the curvature value of the mutation point, searching the optimal clustering number from different preset clustering numbers, and then carrying out clustering analysis on the optimal clustering number based on a K-means clustering algorithm.

10. An identification device for transition areas on urban and rural natural landscape gradients, which is characterized by comprising a memory and a processor, wherein the memory stores the identification method for transition areas on urban and rural natural landscape gradients according to any one of claims 1 to 9, and the processor is used for calling the identification method for transition areas on urban and rural natural landscape gradients stored in the memory to identify the transition areas.