Supplied land idle risk assessment method and system based on spatio-temporal data

Abstract

The invention discloses a supplied land idle risk assessment method and system based on spatio-temporal data, and belongs to the technical field of geographic information, and the method comprises the steps: screening out a first land parcel from supplied land parcels according to the static data of each supplied land parcel; calculating an idle risk assessment index corresponding to the first land parcel according to the static data; extracting historical construction time-space sequence data from the space remote sensing data of the first land parcel from the time when the first land parcel is supplied to now; extracting space-time trend change characteristic data through a time window moving average method by adopting a long-short-term memory network model and taking the historical construction space-time sequence data as input, and predicting a development progress corresponding to the first land parcel according to the space-time trend change characteristic data; and when the development progress and the idle risk assessment index do not meet the preset conditions, determining that the first plot has an idle risk. Through the method and the device, the problem of low timeliness and accuracy of supplied land idle risk assessment can be solved.

Description

Technical Field

[0001] This application relates to the field of geographic information technology, and in particular to a method and system for assessing the risk of idle land already supplied based on spatiotemporal data. Background Technology

[0002] As an important carrier of national economic development, land resources become idle if a supplied land plot has not started construction a year after the scheduled start date, or if construction has started but the area of ​​land used for development and construction is less than one-third of the total area of ​​land to be developed and constructed. Therefore, it is necessary to assess the risk of idle land to avoid wasting land resources.

[0003] Current methods for assessing land idling risk primarily rely on traditional, periodic manual inspections and analysis of static, multi-source data. These methods suffer from significant drawbacks, such as data update delays and a lack of early warning mechanisms. For example, some methods analyze static information from land transfer contracts and irregular reports from municipal and county natural resources departments regarding land development to determine whether supplied land is at risk of idling. However, these methods fail to reflect the timely and objective progress of land development and utilization, and the accuracy of some data is heavily influenced by third parties. Therefore, they are relatively passive and cannot form an effective risk early warning system.

[0004] Therefore, ensuring the timeliness and accuracy of risk assessments for the idleness of supplied land has become a technical problem that needs to be solved. Summary of the Invention

[0005] This application provides a method and system for assessing the risk of idle land already supplied based on spatiotemporal data, which can solve the problems of low timeliness and accuracy in the risk assessment of idle land already supplied in the prior art.

[0006] One embodiment of this application provides a method for assessing the risk of idle land already supplied based on spatiotemporal data, including:

[0007] Collect static data of each supplied land parcel, and select a number of first land parcels from the supplied land parcels based on the static data;

[0008] Based on the static data, calculate the idle risk assessment index corresponding to the first plot of land;

[0009] Using remote sensing technology, spatial remote sensing data of the first plot of land from the time it was supplied to the present is collected, and historical construction spatiotemporal sequence data is extracted from the spatial remote sensing data.

[0010] Using a long short-term memory network model, with the historical construction spatiotemporal sequence data as input, the spatiotemporal trend change feature data is extracted by the time window moving average method, and the development progress of the corresponding first plot is predicted based on the spatiotemporal trend change feature data.

[0011] When neither the development progress nor the idle risk assessment index meets the preset conditions, it is determined that the first plot of land has an idle risk.

[0012] Compared to existing technologies, the above embodiments have the following beneficial effects: Since each indicator in the static data reflects the potential risk characteristics of the land parcel from different dimensions, the static data can be used to initially screen the supplied land parcels and identify the first land parcel that may have risks. Based on the initial assessment of the first land parcel, remote sensing technology is used to enhance the monitoring of the first land parcel, collect dynamically changing spatial remote sensing data, and extract historical construction spatiotemporal sequence data that dynamically reflects the development progress pattern of the first land parcel in both time and space dimensions from the spatial remote sensing data, thereby predicting the development progress of the land parcel. Finally, the idle risk of the first land parcel is comprehensively determined by combining the development progress and the idle risk assessment index, thereby achieving the organic integration of static and dynamic data, compensating for the one-sidedness of a single data source, and improving the accuracy of assessing the idle risk of supplied land.

[0013] Furthermore, the extraction of historical construction spatiotemporal sequence data from the spatial remote sensing data includes:

[0014] The spatial remote sensing data includes: multiple remote sensing images of each of the first plots and the acquisition time of each remote sensing image;

[0015] Using a pre-defined convolutional neural network model, the corresponding local spatial features are extracted from the remote sensing images.

[0016] The commencement time for the first plot of land is determined based on all the local spatial features corresponding to the first plot of land.

[0017] The difference between the collection time and the commencement time is used as a time feature, and combined with the spatial local features, historical spatiotemporal sequence data of the commencement of construction corresponding to each of the first plots are obtained.

[0018] Compared with existing technologies, the above embodiments have the following beneficial effects: By acquiring remote sensing images of the first plot of land, the changes in ground features of each plot of land are identified based on the remote sensing images, thereby avoiding the use of manual periodic inspections and improving the real-time performance of the risk prediction process; furthermore, by extracting spatial local features from the remote sensing images, the commencement time of the plot of land is accurately identified based on the differences in changes of spatial local features, improving the accuracy of subsequent time features; finally, based on the acquisition time of each remote sensing image and the commencement time, the time features corresponding to each spatial local feature are determined, forming a spatiotemporally aligned historical construction spatiotemporal sequence data, thereby improving the accuracy of the output results of the subsequent long short-term memory network model.

[0019] Further, determining the commencement time of construction corresponding to the first plot of land based on all spatial local features of the first plot includes:

[0020] Sort all remote sensing images corresponding to the first plot of land according to the acquisition time;

[0021] Calculate the first Euclidean distance between the spatial local features corresponding to two adjacent remote sensing images in sequence;

[0022] When the first Euclidean distance exceeds the first preset threshold, the acquisition time corresponding to the second remote sensing image of the two remote sensing images is taken as the start time.

[0023] Compared with the prior art, the above embodiments have the following beneficial effects: the remote sensing images are sorted according to the acquisition time, and the Euclidean distance between the corresponding spatial local features of two adjacent remote sensing images is calculated. If the Euclidean distance is too large, it means that the first plot has undergone a large change in surface features. Therefore, the acquisition time of the remote sensing image with a large change in surface features is taken as the start time, thereby improving the accuracy of subsequent time features.

[0024] Further, the step of selecting several first plots from the supplied plots based on the static data includes:

[0025] The static data includes: transfer contract data, infrastructure geographic data, and corresponding enterprise profile data of the users;

[0026] Using a text recognition method, time data is extracted from the land transfer contract data, and several second land parcels are selected from each of the supplied land parcels based on the time data.

[0027] Based on the infrastructure geographic data of the second plot, calculate the second Euclidean distance between the center of the second plot and each infrastructure, and based on each second Euclidean distance, select several third plots from the second plot;

[0028] By using logistic regression and combining the enterprise profile data of the users corresponding to the third land parcel, the credit dimension data of the users is obtained, and a number of the first land parcels are selected from each of the third land parcels based on the credit dimension data.

[0029] Compared to existing technologies, the above embodiments have the following beneficial effects: Since the land transfer contract specifies data at various time points in the land development and utilization process, by extracting this data, second plots with time risks can be screened from the supplied land plots. Furthermore, since the infrastructure geographic data includes the spatial location data of the infrastructure required for land development and utilization, third plots with worse construction conditions can be screened from the second plots based on this infrastructure spatial location data. Finally, based on the corporate profile of the third plot user, first plots corresponding to users with weaker corporate strength or credit can be screened from the third plots. By using static data and performing the above steps to initially screen supplied land plots, the efficiency of determining whether there are plots with idle risk is improved.

[0030] Further, the step of selecting several second plots from the supplied plots based on the time data includes:

[0031] The time data includes: the agreed start date;

[0032] Obtain the difference between the agreed-upon commencement time and the current time.

[0033] The difference is normalized, and the time dimension data of the corresponding supplied land parcel is obtained based on the normalized difference.

[0034] Based on the time dimension data, several second plots are selected from the already supplied plots.

[0035] Compared to existing technologies, the above embodiments have the following beneficial effects: If the current time point is before the agreed commencement time point, the closer the current time point is to the agreed commencement time point, the greater the risk that the current plot of land cannot start construction on schedule; if the current time point is after the agreed commencement time point, the further the current time point is from the agreed commencement time point, and the stipulated development progress has not yet been completed, the greater the risk that the current plot of land will become an idle plot. The greater the risk, the more necessary it is to monitor the plot of land in real time using remote sensing technology. Therefore, the time dimension data calculated using the agreed commencement time point can improve the accuracy of subsequent selection of second plots. Furthermore, since different plots correspond to different agreed commencement time points stipulated in their land transfer contracts, normalization processing ensures the comparability of the time dimension data between the different supplied plots obtained in the end.

[0036] Further, the step of selecting several third plots from the second plots based on each of the second Euclidean distances includes:

[0037] The second Euclidean distance is normalized based on the area of ​​the second plot.

[0038] The second weight corresponding to the second Euclidean distance of the second plot is determined by the entropy method.

[0039] The spatial dimension data of the second plot is obtained based on the normalized second Euclidean distance and the second weight;

[0040] Based on the spatial dimension data, several third plots are selected from the second plot.

[0041] Compared with existing technologies, the above embodiments have the following advantages: First, the Euclidean distance between different infrastructures and the center of the plot is calculated, and then the entropy method is used to objectively allocate the weight of the proximity of infrastructures to distinguish the importance of different infrastructures to the development progress of the plot; further, normalization processing is used to eliminate the interference of the development area corresponding to different plots on the subsequent spatial dimension data calculation, ensuring the effectiveness of the subsequent screening process.

[0042] Further, the step of calculating the idle risk assessment index corresponding to the first land parcel based on the static data includes:

[0043] Based on the data combinations corresponding to each of the first plots, a positive ideal solution and a negative ideal solution are constructed; the data combinations include: the credit dimension data, the time dimension data, and the spatial dimension data;

[0044] Calculate the third Euclidean distance between the data combination corresponding to the first plot and the positive ideal solution, and the fourth Euclidean distance between the data combination and the negative ideal solution, respectively.

[0045] The idle risk assessment index corresponding to the first plot of land is calculated based on the third Euclidean distance and the fourth Euclidean distance.

[0046] Compared with the prior art, the above embodiments have the following beneficial effects: After the first plot of land is selected, based on the data of the first plot of land in terms of time, space and corporate credit, the difference in idle risk of each first plot of land is objectively evaluated through the ideal solution method, so as to achieve an objective ranking of the idle risk level of different first plots of land. This provides a quantitative basis for the subsequent comprehensive evaluation of the actual idle risk of the first plot of land by combining dynamic data, and improves the accuracy of determining the final idle risk plots.

[0047] Another embodiment of this application provides a risk assessment system for the idleness of supplied land based on spatiotemporal data, including: a land parcel screening module, an idleness risk assessment index calculation module, a spatiotemporal sequence data extraction module, a development progress prediction module, and an idleness risk determination module;

[0048] The land parcel screening module collects static data of each supplied land parcel and selects several first land parcels from the supplied land parcels based on the static data.

[0049] The idle risk assessment index calculation module is used to calculate the idle risk assessment index corresponding to the first plot of land based on the static data.

[0050] The spatiotemporal sequence data extraction module is used to collect spatial remote sensing data of the first plot from the time it was supplied to the present using remote sensing technology, and to extract historical construction spatiotemporal sequence data from the spatial remote sensing data.

[0051] The development progress prediction module is used to employ a long short-term memory network model, take the historical construction spatiotemporal sequence data as input, extract spatiotemporal trend change feature data through a time window moving average method, and predict the development progress of the corresponding first plot based on the spatiotemporal trend change feature data.

[0052] The idle risk determination module is used to determine that the first plot of land has an idle risk when neither the development progress nor the idle risk assessment index meets the preset conditions.

[0053] Furthermore, the spatiotemporal sequence data extraction module includes: a spatial local feature extraction unit, a construction start time acquisition unit, and a time feature extraction unit; the spatiotemporal sequence data extraction module is used to extract historical construction start spatiotemporal sequence data from the spatial remote sensing data, specifically:

[0054] The spatial remote sensing data includes: multiple remote sensing images of each of the first plots and the acquisition time of each remote sensing image;

[0055] The spatial local feature extraction unit is used to extract corresponding spatial local features from the remote sensing image using a preset convolutional neural network model.

[0056] The construction start time acquisition unit is used to determine the construction start time corresponding to the first plot of land based on all spatial local features corresponding to the first plot of land.

[0057] The time feature extraction unit is used to take the difference between the collection time and the commencement time as a time feature, and combine it with the spatial local features to obtain the historical spatiotemporal sequence data of each first plot of land.

[0058] Further, the land parcel screening module includes: a second land parcel screening unit, a third land parcel screening unit, and a first land parcel screening unit; the land parcel screening module is used to screen out a number of first land parcels from the supplied land parcels based on the static data, specifically:

[0059] The static data includes: transfer contract data, infrastructure geographic data, and corresponding enterprise profile data of the users;

[0060] The second land parcel screening unit is used to extract time data from the land transfer contract data using a text recognition method, and to screen a number of second land parcels from each of the supplied land parcels based on the time data;

[0061] The third land parcel screening unit is used to calculate the second Euclidean distance between the center of the second land parcel and each infrastructure based on the infrastructure geographic data of the second land parcel, and to screen a number of third land parcels from the second land parcel based on each second Euclidean distance;

[0062] The first land parcel screening unit is used to obtain the credit dimension data of the user by combining the enterprise profile data of the user corresponding to the third land parcel with the logistic regression method, and to screen a number of the first land parcels from each of the third land parcels according to the credit dimension data. Attached Figure Description

[0063] To more clearly illustrate the technical solution of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0064] Figure 1 This is a flowchart illustrating a method for assessing the risk of idle land supplied based on spatiotemporal data, provided in some embodiments of this application.

[0065] Figure 2 This is a schematic diagram of the structure of a spatiotemporal data-based risk assessment system for the idleness of supplied land, provided in some embodiments of this application. Detailed Implementation

[0066] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0067] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the application; the terms “comprising” and “having”, and any variations thereof, in the specification, claims, and foregoing description of the drawings are intended to cover non-exclusive inclusion.

[0068] In the description of the embodiments of this application, technical terms such as "first" and "second" are used only to distinguish different objects and should not be construed as indicating or implying relative importance or implicitly specifying the number, specific order, or primary and secondary relationship of the indicated technical features. In the description of the embodiments of this application, "multiple" means two or more, unless otherwise explicitly defined.

[0069] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0070] In the description of the embodiments in this application, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this document generally indicates that the preceding and following related objects have an "or" relationship.

[0071] In the description of the embodiments of this application, the term "multiple" refers to two or more (including two), similarly, "multiple sets" refers to two or more (including two sets), and "multiple pieces" refers to two or more (including two pieces).

[0072] In the description of the embodiments of this application, unless otherwise expressly specified and limited, technical terms such as "installation," "connection," "joining," and "fixing" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. For those skilled in the art, the specific meaning of the above terms in the embodiments of this application can be understood according to the specific circumstances.

[0073] Current methods for assessing land idling risk primarily rely on traditional, periodic manual inspections and analysis of static, multi-source data. These methods suffer from significant drawbacks, such as data update delays and a lack of early warning mechanisms. For example, some methods analyze static information from land transfer contracts and irregular reports from municipal and county natural resources departments regarding land development to determine whether supplied land is at risk of idling. However, these methods fail to reflect the timely and objective progress of land development and utilization, and the accuracy of some data is heavily influenced by third parties. Therefore, they are relatively passive and cannot form an effective risk early warning system.

[0074] Therefore, ensuring the timeliness and accuracy of assessing the risk of idle land that has already been supplied has become a technical problem that needs to be solved.

[0075] Please refer to Figure 1 To address the issues of low timeliness and accuracy in the risk assessment of idle land already supplied in existing technologies, this application provides a method for risk assessment of idle land already supplied based on spatiotemporal data, comprising steps S101 to S105, specifically:

[0076] S101: Collect static data of each supplied land parcel, and select a number of first land parcels from the supplied land parcels based on the static data.

[0077] Furthermore, in some embodiments of this application, the static data includes: transfer contract data, infrastructure geographic data, and corresponding user enterprise profile data.

[0078] Furthermore, in some embodiments of this application, the collection of static data includes: collecting text data of the transfer contracts for each supplied land parcel from the land supply database as transfer contract data; collecting the first geographic coordinates of infrastructure within a preset range of each supplied land parcel from the geographic database as geographic data, wherein the infrastructure includes, but is not limited to, transportation facilities, water supply and drainage facilities, and energy facilities, the aforementioned infrastructure being used to ensure access, water supply, and electricity supply needs during the development process of each land parcel; and collecting enterprise profile data of the corresponding users of each supplied land parcel through an enterprise information interface, the enterprise profile data including, but not limited to, the user's registered capital size information, historical default record information, and industry reputation index information, etc.

[0079] Furthermore, in some embodiments of this application, the step of selecting a plurality of first land parcels from the supplied land parcels based on the static data includes:

[0080] Using a text recognition method, time data is extracted from the land transfer contract data, and several second land parcels are selected from each of the supplied land parcels based on the time data.

[0081] Based on the infrastructure geographic data of the second plot, calculate the second Euclidean distance between the center of the second plot and each infrastructure, and based on each second Euclidean distance, select several third plots from the second plot;

[0082] By using logistic regression and combining the enterprise profile data of the users corresponding to the third land parcel, the credit dimension data of the users is obtained, and a number of the first land parcels are selected from each of the third land parcels based on the credit dimension data.

[0083] Furthermore, in some embodiments of this application, the step of extracting time data from the land transfer contract data using a text recognition method includes: obtaining land transfer contract data for the target land parcel from a land supply database, parsing the timestamps and timestamp descriptions in the land transfer contract data, and obtaining the time data of the land transfer contract. It is understood that each land transfer contract contains multiple timestamps, thus allowing the acquisition of multiple agreed time nodes. The meaning of each agreed time node can be determined by combining the timestamp descriptions. For example, timestamps may include: agreed land delivery time and agreed construction commencement time. For instance, by extracting the agreed land delivery time, if the land is not delivered by the agreed delivery time, it means the company may not be able to commence development by the agreed construction commencement time; conversely, if the company fails to commence construction within the agreed construction commencement time, it also means the company cannot carry out land development as scheduled within the stipulated time. The correspondence between each extracted agreed time node and its corresponding timestamp can be understood as follows: if the land transfer contract for a certain plot stipulates that the commencement date is June 1, 2023, then "agreed commencement date" is the timestamp description, and "June 1, 2023" is the timestamp.

[0084] Furthermore, in some embodiments of this application, the calculation of the second Euclidean distance between the center of the second plot and each infrastructure includes: obtaining the second geographic coordinates of the center of each second plot, and calculating the Euclidean distance between the second geographic coordinates and the first geographic coordinates of each infrastructure corresponding to the second plot.

[0085] Furthermore, in some embodiments of this application, the step of obtaining the user's credit dimension data by combining the enterprise profile data of the user corresponding to the third land parcel includes: collecting enterprise profile data through an enterprise information interface, wherein the enterprise profile data includes: the default records of the user corresponding to the supplied land parcel in previous land development (such as whether construction started on schedule and whether the user corresponding to the supplied land parcel currently has idle land that has been supplied but not used), the enterprise qualifications, enterprise cash flow data, enterprise historical land development record data, and enterprise registration scale data of the user corresponding to each supplied land parcel, etc. Based on the above enterprise profile data, a credit rating score is calculated using logistic regression to obtain the user's credit dimension data. For example, the user of a certain land parcel is a real estate company with a registered capital of 50 million yuan, no historical default records, and an industry reputation index of 85 points. Using logistic regression, this information is converted into a credit rating score. Assuming that the standardized registered capital is 0.6, the default record is 0, and the reputation index is 0.8, the credit dimension data is obtained by weighted calculation as 0.75. This application does not impose specific restrictions on the standardized scoring method used in the calculation of credit dimension data.

[0086] As can be seen from the above embodiments, since the land transfer contract specifies data at various time points in the land development and utilization process, by extracting this data, a second plot of land with time risks can be screened from the supplied land plots. Furthermore, since the infrastructure geographic data includes the spatial location data of the infrastructure required for the land development and utilization, a third plot with worse construction conditions can be screened from the second plot. Finally, based on the corporate profile of the third plot's user, a first plot corresponding to a user with weaker corporate strength or credit can be screened from the third plot. By using static data and performing the above steps to initially screen the supplied land plots, the efficiency of determining whether there are plots with idle risk is improved.

[0087] Furthermore, in some embodiments of this application, the step of selecting several second plots from the supplied plots based on the time data includes:

[0088] The time data includes: the agreed start date;

[0089] Obtain the difference between the agreed-upon commencement time and the current time.

[0090] The difference is normalized, and the time dimension data of the corresponding supplied land parcel is obtained based on the normalized difference.

[0091] Based on the time dimension data, several second plots are selected from the already supplied plots.

[0092] Furthermore, in some embodiments of this application, the normalization process for the difference includes: obtaining the total agreed construction duration in the land transfer contract corresponding to the difference; and dividing the difference by the total agreed construction duration to obtain the normalized data.

[0093] Furthermore, in some embodiments of this application, obtaining the time dimension data corresponding to the supplied land parcel based on the normalized difference includes: using the normalized difference as the time dimension data.

[0094] As can be seen from the above examples, if the current time point is before the agreed commencement date, the closer the current time point is to the agreed commencement date, the greater the risk that the current plot of land will not be able to start construction on schedule. Conversely, if the current time point is after the agreed commencement date, the further the current time point is from the agreed commencement date, the greater the risk that the current plot of land will not be able to complete land development on time. The greater the risk, the more necessary it is to monitor the plot of land in real time using remote sensing technology. Therefore, the time dimension data calculated using the agreed commencement date can improve the accuracy of subsequent selection of second plots. Furthermore, since the total commencement time stipulated in the land transfer contracts differs for different plots, normalization processing ensures the comparability of the time dimension data between the different supplied plots.

[0095] Furthermore, in some embodiments of this application, the step of selecting a plurality of third plots from the second plots based on each of the second Euclidean distances includes:

[0096] The second Euclidean distance is normalized based on the area of ​​the second plot.

[0097] The second weight corresponding to the second Euclidean distance of the second plot is determined by the entropy method.

[0098] The spatial dimension data of the second plot is obtained based on the normalized second Euclidean distance and the second weight;

[0099] Based on the spatial dimension data, several third plots are selected from the second plot.

[0100] Furthermore, in some embodiments of this application, the step of normalizing the second Euclidean distance based on the area of ​​the second plot includes: dividing the second Euclidean distance by the total area of ​​the corresponding second plot to obtain the normalized second Euclidean distance.

[0101] Furthermore, in some embodiments of this application, determining the second weight corresponding to the second Euclidean distance of the second land parcel using the entropy method includes:

[0102] The second Euclidean distance between the same type of infrastructure and each second plot is selected, and the third weight of this type of infrastructure is determined by the following formula:

[0103]

[0104] Among them, H j p is the third weight for the j-th type of infrastructure; ij Let be the second Euclidean distance between the i-th second plot and the j-th type of infrastructure.

[0105] The second weight is determined by the third weight of the infrastructure corresponding to the second Euclidean distance of the second plot. For example, if the third weight of the current road facilities is 0.4, then the second weight of all the second Euclidean distances related to the road facilities is also set to 0.4.

[0106] Furthermore, in some embodiments of this application, obtaining the spatial dimension data of the corresponding second plot based on the normalized second Euclidean distance and the second weight includes: multiplying each normalized second Euclidean distance by the corresponding second weight and then summing them to obtain the spatial dimension data of the corresponding second plot.

[0107] As can be seen from the above embodiments, this application first calculates the Euclidean distance between different infrastructures and the center of the plot, and then uses the entropy method to objectively allocate the weight of the infrastructure proximity to distinguish the importance of different infrastructures to the development progress of the plot; further, it eliminates the interference of the development area corresponding to different plots on the subsequent spatial dimension data calculation through normalization processing, so as to ensure the effectiveness of the subsequent screening process.

[0108] S102: Calculate the idle risk assessment index corresponding to the first plot of land based on the static data.

[0109] Furthermore, in some embodiments of this application, the step of calculating the idle risk assessment index corresponding to the first land parcel based on the static data includes:

[0110] Based on the data combinations corresponding to each of the first plots, a positive ideal solution and a negative ideal solution are constructed; the data combinations include: the credit dimension data, the time dimension data, and the spatial dimension data;

[0111] Calculate the third Euclidean distance between the data combination corresponding to the first plot and the positive ideal solution, and the fourth Euclidean distance between the data combination and the negative ideal solution, respectively.

[0112] The idle risk assessment index corresponding to the first plot of land is calculated based on the third Euclidean distance and the fourth Euclidean distance.

[0113] Furthermore, in some embodiments of this application, the step of constructing positive and negative ideal solutions based on the data combinations corresponding to each of the first land parcels includes:

[0114] The positive ideal solution is obtained by finding the maximum value of each dimension in all data combinations of the first plot. For example, if the data combinations of the first plot consist of (1, 1.2, 1.5), (0.7, 1, 2), and (0.8, 0.4, 1), then the positive ideal solution is (1, 1.2, 2). Similarly, the negative ideal solution is obtained by finding the minimum value of each dimension in all data combinations of the first plot. For example, the negative ideal solution in the above example is (0.7, 0.4, 1).

[0115] Furthermore, in some embodiments of this application, the step of calculating the idle risk assessment index corresponding to the first plot based on the third Euclidean distance and the fourth Euclidean distance includes: using the ratio of the third Euclidean distance to the fourth Euclidean distance as the idle risk assessment index.

[0116] As can be seen from the above embodiments, after selecting the first plot of land, this application objectively evaluates the differences in idle risk of each first plot of land based on the data of the first plot of land in terms of time, space and corporate credit through an ideal solution, thereby achieving an objective ranking of the idle risk levels of different first plots of land. This provides a quantitative basis for the subsequent comprehensive evaluation of the actual idle risk of the first plot of land by combining dynamic data, and improves the accuracy of determining the final idle risk plots of land.

[0117] S103: Using remote sensing technology, collect spatial remote sensing data of the first plot of land from the time it was supplied to the present, and extract historical construction spatiotemporal sequence data from the spatial remote sensing data.

[0118] Furthermore, in some embodiments of this application, the step of extracting historical construction spatiotemporal sequence data from the spatial remote sensing data includes:

[0119] The spatial remote sensing data includes: multiple remote sensing images of each of the first plots and the acquisition time of each remote sensing image;

[0120] Using a pre-defined convolutional neural network model, the corresponding local spatial features are extracted from the remote sensing images.

[0121] The commencement time for the first plot of land is determined based on all the local spatial features corresponding to the first plot of land.

[0122] The difference between the collection time and the commencement time is used as a time feature, and combined with the spatial local features, historical spatiotemporal sequence data of the commencement of construction corresponding to each of the first plots are obtained.

[0123] Furthermore, in some embodiments of this application, the step of extracting corresponding spatial local features from the remote sensing image using a preset convolutional neural network model includes:

[0124] A pre-trained remote sensing image classification model (such as Sentinel-2ResNet50) is used. The input is a single remote sensing image, and the output is a spatial local feature map. The spatial local feature map is generated as follows: texture, shape and other features (such as road lines and building outlines) are extracted from the remote sensing image through convolutional layers; the extracted features are further reduced in dimensionality and key spatial information is preserved through pooling layers; finally, a fixed-length feature vector is generated through global average pooling.

[0125] By acquiring remote sensing images of the first plot of land, and identifying changes in ground features for each plot based on these images, the need for regular manual inspections is eliminated, improving the real-time nature of the risk prediction process. Furthermore, by extracting spatial local features from the remote sensing images, the commencement time of each plot is accurately identified based on the differences in these features, improving the accuracy of subsequent temporal features. Finally, based on the acquisition time of each remote sensing image and the commencement time, the temporal features corresponding to each spatial local feature are determined, forming a spatiotemporally aligned historical construction commencement spatiotemporal sequence data, thereby improving the accuracy of the output results of the subsequent Long Short-Term Memory Network model.

[0126] Furthermore, in some embodiments of this application, determining the commencement time of construction corresponding to the first plot of land based on all spatial local features corresponding to the first plot includes:

[0127] Sort all remote sensing images corresponding to the first plot of land according to the acquisition time;

[0128] Calculate the first Euclidean distance between the spatial local features corresponding to two adjacent remote sensing images in sequence;

[0129] When the first Euclidean distance exceeds the first preset threshold, the acquisition time corresponding to the second remote sensing image of the two remote sensing images is taken as the start time.

[0130] Furthermore, in some embodiments of this application, when the first Euclidean distance exceeds a first preset threshold, the acquisition time corresponding to the latter of the two remote sensing images is taken as the start-up time, including:

[0131] Assuming that the spatial local feature vectors of two adjacent remote sensing images are F1 and F2, where F1 is the spatial local feature vector of the first acquired remote sensing image, then the first Euclidean distance d = ||F1-F2|| is calculated. If d > the first preset value, then the acquisition time of the remote sensing image corresponding to F2 is determined to be the start time.

[0132] As can be seen from the above embodiments, this application sorts remote sensing images according to the acquisition time and calculates the Euclidean distance between corresponding spatial local features of two adjacent remote sensing images. If the Euclidean distance is too large, it means that the first plot has undergone a large change in surface features. Therefore, the acquisition time of the remote sensing image with a large change in surface features is taken as the start time, thereby improving the accuracy of subsequent time features.

[0133] S104: Using a long short-term memory network model, with the historical construction spatiotemporal sequence data as input, the spatiotemporal trend change feature data is extracted by the time window moving average method, and the development progress of the corresponding first plot is predicted based on the spatiotemporal trend change feature data.

[0134] Furthermore, in some embodiments of this application, the Long Short-Term Memory network model is a special recurrent neural network that can learn long-term dependencies, effectively process sequence data, and predict future results based on sequence data.

[0135] Furthermore, in some embodiments of this application, the development progress refers to the land development progress one year after the currently predicted land parcel reaches the agreed commencement date.

[0136] S105: When neither the development progress nor the idle risk assessment index meets the preset conditions, it is determined that the first plot of land has an idle risk.

[0137] Furthermore, in some embodiments of this application, the development progress and the idle risk assessment index do not meet the preset conditions, including: the development progress is less than one-third of the total area of ​​land to be developed and constructed, and the idle risk assessment index is greater than a preset threshold. This application does not limit the threshold.

[0138] In summary, the method for assessing the risk of idle land supplied based on spatiotemporal data provided in this application has the following beneficial effects: Since each indicator in the static data reflects the potential risk characteristics of the land parcel from different dimensions, the static data can be used to initially screen supplied land parcels and identify the first land parcel that may have risks. Based on the initial assessment of the first land parcel, remote sensing technology is used to strengthen the monitoring of the first land parcel, collect dynamically changing spatial remote sensing data, and extract historical spatiotemporal sequence data of construction commencement that dynamically reflects the development progress pattern of the first land parcel in both time and space dimensions from the spatial remote sensing data, thereby predicting the development progress of the land parcel. Finally, the idle risk of the first land parcel is comprehensively determined by combining the development progress and the idle risk assessment index, thereby achieving the organic integration of static and dynamic data, compensating for the one-sidedness of a single data source, and improving the accuracy of assessing the idle risk of supplied land.

[0139] like Figure 2 As shown, based on the above-mentioned method embodiments, an embodiment of this application provides a land idling risk assessment system based on spatiotemporal data, including: a land parcel screening module 201, an idling risk assessment index calculation module 202, a spatiotemporal sequence data extraction module 203, a development progress prediction module 204, and an idling risk determination module 205.

[0140] Further, in some embodiments of this application, the land parcel screening module 201 collects static data of each supplied land parcel and selects several first land parcels from the supplied land parcels based on the static data; the idle risk assessment index calculation module 202 is used to calculate the idle risk assessment index corresponding to the first land parcel based on the static data; the spatiotemporal sequence data extraction module 203 is used to collect spatial remote sensing data of the first land parcel from the time it was supplied to the present using remote sensing technology, and extract historical construction spatiotemporal sequence data from the spatial remote sensing data; the development progress prediction module 204 is used to use a long short-term memory network model, with the historical construction spatiotemporal sequence data as input, extract spatiotemporal trend change feature data through a time window moving average method, and predict the development progress of the corresponding first land parcel based on the spatiotemporal trend change feature data; the idle risk determination module 205 is used to determine that the first land parcel has idle risk when neither the development progress nor the idle risk assessment index meets preset conditions.

[0141] Further, in some embodiments of this application, the spatiotemporal sequence data extraction module 203 includes: a spatial local feature extraction unit, a construction start time acquisition unit, and a time feature extraction unit; the spatiotemporal sequence data extraction module 203 is used to extract historical construction spatiotemporal sequence data from the spatial remote sensing data, specifically: the spatial remote sensing data includes: multiple remote sensing images of each first plot and the acquisition time of each remote sensing image; the spatial local feature extraction unit is used to extract corresponding spatial local features from the remote sensing images using a preset convolutional neural network model; the construction start time acquisition unit is used to determine the construction start time of the corresponding first plot based on all spatial local features corresponding to the first plot; the time feature extraction unit is used to use the difference between the acquisition time and the construction start time as a time feature, combined with the spatial local features, to obtain historical construction spatiotemporal sequence data corresponding to each first plot.

[0142] Furthermore, in some embodiments of this application, the construction start time acquisition unit is used to determine the construction start time corresponding to the first plot of land based on all spatial local features corresponding to the first plot of land, including: sorting all remote sensing images corresponding to the first plot of land according to the acquisition time; sequentially calculating the first Euclidean distance between the spatial local features corresponding to two adjacent remote sensing images; when the first Euclidean distance exceeds a first preset threshold, taking the acquisition time corresponding to the latter remote sensing image of the two remote sensing images as the construction start time.

[0143] Further, in some embodiments of this application, the land parcel screening module 201 includes: a second land parcel screening unit, a third land parcel screening unit, and a first land parcel screening unit; the land parcel screening module 201 is used to screen out a plurality of first land parcels from the supplied land parcels based on the static data, specifically: the static data includes: land transfer contract data, infrastructure geographic data, and corresponding user enterprise profile data; the second land parcel screening unit is used to extract time data from the land transfer contract data using a text recognition method, and screen a plurality of second land parcels from each of the supplied land parcels based on the time data; the third land parcel screening unit is used to calculate the second Euclidean distance between the center of the second land parcel and each infrastructure based on the infrastructure geographic data of the second land parcel, and screen a plurality of third land parcels from the second land parcels based on each of the second Euclidean distances; the first land parcel screening unit is used to obtain the user's credit dimension data by combining the user's enterprise profile data corresponding to the third land parcel using a logistic regression method, and screen a plurality of first land parcels from each of the third land parcels based on the credit dimension data.

[0144] Furthermore, in some embodiments of this application, the second land parcel screening unit is used to screen a plurality of second land parcels from each of the supplied land parcels based on the time data, including: wherein the time data includes: an agreed construction start time; obtaining the difference between the agreed construction start time and the current time; normalizing the difference and obtaining the time dimension data of the corresponding supplied land parcel based on the normalized difference; and screening a plurality of second land parcels from the supplied land parcels based on the time dimension data.

[0145] Furthermore, in some embodiments of this application, the third land parcel screening unit is used to screen a plurality of third land parcels from the second land parcels based on each of the second Euclidean distances, including: normalizing the second Euclidean distances based on the area of ​​the second land parcels; determining a second weight for the second land parcels corresponding to the second Euclidean distances using the entropy method; obtaining spatial dimension data corresponding to the second land parcels based on the normalized second Euclidean distances and the second weights; and screening a plurality of third land parcels from the second land parcels based on the spatial dimension data.

[0146] Further, in some embodiments of this application, the idle risk assessment index calculation module 202 includes: an ideal solution construction unit, an Euclidean distance calculation unit, and an idle risk assessment index calculation unit; the idle risk assessment index calculation module 202 is used to calculate the idle risk assessment index corresponding to the first land parcel based on the static data, including: the ideal solution construction unit is used to construct a positive ideal solution and a negative ideal solution based on the data combination corresponding to each first land parcel; the data combination includes: the credit dimension data, the time dimension data, and the spatial dimension data; the Euclidean distance calculation unit is used to calculate the third Euclidean distance between the data combination corresponding to the first land parcel and the positive ideal solution and the fourth Euclidean distance between the data combination and the negative ideal solution, respectively; the idle risk assessment index calculation unit is used to calculate the idle risk assessment index corresponding to the first land parcel based on the third Euclidean distance and the fourth Euclidean distance.

[0147] It is understood that the above system item embodiments correspond to the method item embodiments of this application, and can implement the method for assessing the risk of idle land supplied based on spatiotemporal data provided by any of the above method item embodiments of this application.

[0148] It should be noted that the system embodiments described above are merely illustrative, and some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Furthermore, in the accompanying drawings of the system embodiments provided in this application, the connection relationships between modules indicate that they have communication connections, which can be specifically implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement this without any creative effort.

[0149] Based on the above embodiments of the method for assessing the risk of idle land supplied based on spatiotemporal data, another embodiment of this application provides a terminal device, which includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the method for assessing the risk of idle land supplied based on spatiotemporal data of any embodiment of this application.

[0150] For example, in this embodiment, the computer program can be divided into one or more modules, which are stored in the memory and executed by the processor to complete this application. The one or more module units may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in the terminal device.

[0151] The terminal device may be a desktop computer, laptop, handheld computer, or cloud server, etc. The terminal device may include, but is not limited to, a processor and a memory.

[0152] The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor. The processor is the control center of the terminal device, connecting all parts of the terminal device via various interfaces and lines.

[0153] Based on the above-described method embodiments, another embodiment of this application provides a computer-readable storage medium including a stored computer program, wherein, when the computer program is running, it controls the device where the computer-readable storage medium is located to execute the method for assessing the risk of idle supplied land based on spatiotemporal data as described in any of the above-described method embodiments of this application.

[0154] The modules / units integrated in the device / terminal equipment, if implemented as software functional units and sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

Claims

1. A method for assessing the risk of idle land already supplied based on spatiotemporal data, characterized in that, include: Collect static data of each supplied land parcel, and select a number of first land parcels from the supplied land parcels based on the static data; Based on the static data, calculate the idle risk assessment index corresponding to the first plot of land; Using remote sensing technology, spatial remote sensing data of the first plot of land from the time it was supplied to the present is collected, and historical construction spatiotemporal sequence data is extracted from the spatial remote sensing data. Using a long short-term memory network model, with the historical construction spatiotemporal sequence data as input, the spatiotemporal trend change feature data is extracted by the time window moving average method, and the development progress of the corresponding first plot is predicted based on the spatiotemporal trend change feature data. When neither the development progress nor the idle risk assessment index meets the preset conditions, it is determined that the first plot of land has an idle risk.

2. The method for assessing the risk of idle land already supplied based on spatiotemporal data as described in claim 1, characterized in that, The extraction of historical construction spatiotemporal sequence data from the spatial remote sensing data includes: The spatial remote sensing data includes: multiple remote sensing images of each of the first plots and the acquisition time of each remote sensing image; Using a pre-defined convolutional neural network model, the corresponding local spatial features are extracted from the remote sensing images. The commencement time for the first plot of land is determined based on all the local spatial features corresponding to the first plot of land. The difference between the collection time and the commencement time is used as a time feature, and combined with the spatial local features, historical spatiotemporal sequence data of the commencement of construction corresponding to each of the first plots are obtained.

3. The method for assessing the risk of idle land already supplied based on spatiotemporal data as described in claim 2, characterized in that, The step of determining the commencement time of construction corresponding to the first plot of land based on all spatial local features of the first plot includes: Sort all remote sensing images corresponding to the first plot of land according to the acquisition time; Calculate the first Euclidean distance between the spatial local features corresponding to two adjacent remote sensing images in sequence; When the first Euclidean distance exceeds the first preset threshold, the acquisition time corresponding to the second remote sensing image of the two remote sensing images is taken as the start time.

4. The method for assessing the risk of idle land already supplied based on spatiotemporal data as described in claim 1, characterized in that, The step of selecting several first plots from the supplied plots based on the static data includes: The static data includes: transfer contract data, infrastructure geographic data, and corresponding enterprise profile data of the users; Using a text recognition method, time data is extracted from the land transfer contract data, and several second land parcels are selected from each of the supplied land parcels based on the time data. Based on the infrastructure geographic data of the second plot, calculate the second Euclidean distance between the center of the second plot and each infrastructure, and based on each second Euclidean distance, select several third plots from the second plot; By using logistic regression and combining the enterprise profile data of the users corresponding to the third land parcel, the credit dimension data of the users is obtained, and a number of the first land parcels are selected from each of the third land parcels based on the credit dimension data.

5. The method for assessing the risk of idle land already supplied based on spatiotemporal data as described in claim 4, characterized in that, The process involves selecting several second plots from the already supplied plots based on the time data. include: The time data includes: the agreed start date; Obtain the difference between the agreed commencement time and the current time. The difference is normalized, and the time dimension data of the corresponding supplied land parcel is obtained based on the normalized difference. Based on the time dimension data, several second plots are selected from the already supplied plots.

6. The method for assessing the risk of idle land already supplied based on spatiotemporal data as described in claim 5, characterized in that, The step of selecting several third plots from the second plots based on each of the second Euclidean distances includes: The second Euclidean distance is normalized based on the area of ​​the second plot. The second weight corresponding to the second Euclidean distance of the second plot is determined by the entropy method. The spatial dimension data of the second plot is obtained based on the normalized second Euclidean distance and the second weight; Based on the spatial dimension data, several third plots are selected from the second plot.

7. The method for assessing the risk of idle land already supplied based on spatiotemporal data as described in claim 6, characterized in that, The step of calculating the idle risk assessment index corresponding to the first land parcel based on the static data includes: Based on the data combinations corresponding to each of the first plots, a positive ideal solution and a negative ideal solution are constructed; the data combinations include: the credit dimension data, the time dimension data, and the spatial dimension data; Calculate the third Euclidean distance between the data combination corresponding to the first plot and the positive ideal solution, and the fourth Euclidean distance between the data combination and the negative ideal solution, respectively. The idle risk assessment index corresponding to the first plot of land is calculated based on the third Euclidean distance and the fourth Euclidean distance.

8. A system for assessing the risk of idle land already supplied based on spatiotemporal data, characterized in that, include: The module includes a land parcel screening module, an idle risk assessment index calculation module, a spatiotemporal sequence data extraction module, a development progress prediction module, and an idle risk determination module. The land parcel screening module collects static data of each supplied land parcel and selects several first land parcels from the supplied land parcels based on the static data. The idle risk assessment index calculation module is used to calculate the idle risk assessment index corresponding to the first plot of land based on the static data. The spatiotemporal sequence data extraction module is used to collect spatial remote sensing data of the first plot from the time it was supplied to the present using remote sensing technology, and to extract historical construction spatiotemporal sequence data from the spatial remote sensing data. The development progress prediction module is used to employ a long short-term memory network model, take the historical construction spatiotemporal sequence data as input, extract spatiotemporal trend change feature data through a time window moving average method, and predict the development progress of the corresponding first plot based on the spatiotemporal trend change feature data. The idle risk determination module is used to determine that the first plot of land has an idle risk when neither the development progress nor the idle risk assessment index meets the preset conditions.

9. A risk assessment system for the vacancy of supplied land based on spatiotemporal data as described in claim 8, characterized in that, The spatiotemporal sequence data extraction module includes: a spatial local feature extraction unit, a construction start time acquisition unit, and a time feature extraction unit; the spatiotemporal sequence data extraction module is used to extract historical construction start spatiotemporal sequence data from the spatial remote sensing data, specifically: The spatial remote sensing data includes: multiple remote sensing images of each of the first plots and the acquisition time of each remote sensing image; The spatial local feature extraction unit is used to extract corresponding spatial local features from the remote sensing image using a preset convolutional neural network model. The construction start time acquisition unit is used to determine the construction start time corresponding to the first plot of land based on all spatial local features corresponding to the first plot of land. The time feature extraction unit is used to take the difference between the collection time and the commencement time as a time feature, and combine it with the spatial local features to obtain the historical spatiotemporal sequence data of each first plot of land.

10. The system for assessing the risk of idle land supply based on spatiotemporal data as described in claim 8, characterized in that, The land parcel screening module includes: a second land parcel screening unit, a third land parcel screening unit, and a first land parcel screening unit; the land parcel screening module is used to screen out a number of first land parcels from the supplied land parcels based on the static data, specifically as follows: The static data includes: transfer contract data, infrastructure geographic data, and corresponding enterprise profile data of the users; The second land parcel screening unit is used to extract time data from the land transfer contract data using a text recognition method, and to screen a number of second land parcels from each of the supplied land parcels based on the time data; The third land parcel screening unit is used to calculate the second Euclidean distance between the center of the second land parcel and each infrastructure based on the infrastructure geographic data of the second land parcel, and to screen a number of third land parcels from the second land parcel based on each second Euclidean distance; The first land parcel screening unit is used to obtain the credit dimension data of the user by combining the enterprise profile data of the user corresponding to the third land parcel with the logistic regression method, and to screen a number of the first land parcels from each of the third land parcels according to the credit dimension data.

Images