A method, product, medium and device for intelligent management and control of national land space planning
By building a basic geographic information and natural resource management data sub-base for land space planning, using deep learning and knowledge graph technology, combined with policy and regulatory data, intelligent control is achieved, and two-dimensional control is not intuitive and intention-conductive in traditional land space planning. The problems of inaccuracy and low degree of automation have improved the accuracy and efficiency of planning implementation management.
Patent Information
- Application Number
- CN202411280427.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-13
- Publication Date
- 2024-12-03
- Estimated Expiration
- 2044-09-13
AI Technical Summary
In the implementation of traditional land space planning, there are problems such as two-dimensional plane control modes that are not intuitive, inaccurate transmission of text planning intentions, and low degree of automation, resulting in inaccurate and inefficient planning implementation management.
By obtaining basic geographic information data and natural resource management data, building basic geographic information data sub-base and natural resource management data sub-base, using deep learning methods for remote sensing interpretation, combining policy and regulatory data to build a knowledge graph to achieve three-dimensional intelligent control results Visual expression.
It has improved the accuracy and efficiency of intelligent control of land space planning, enhanced the visual expression ability of planning implementation management, and formed a highly automated and intelligent intelligent intelligent land space planning and control brain.
Smart Images

Figure CN118799127B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field of national land space planning, and in particular to a method, product, medium and equipment for intelligent management and control of national land space planning. Background Art
[0002] At present, with the rapid advancement of urbanization, natural resource management is facing new complex environments and tasks. Due to limited manpower, information asymmetry, lack of precise technical software support and other reasons in the management process, the implementation stage of the national land space planning has led to defects in the inaccurate detection, recognition and management of natural resource elements. The existing planning implementation and management work faces three major challenges: First, the traditional two-dimensional plane control mode is not intuitive: traditional planning implementation focuses on the control of two-dimensional planes, but lacks consideration of three-dimensional spatial forms, and often ignores the coordination between local construction and the overall three-dimensional spatial environment, thereby reducing the accuracy of planning implementation and management; Second, the traditional text planning intention is not accurately transmitted: the large scale differences and different responsibilities between the various levels of planning lead to the lack of effective support for the transmission and connection of planning and control content between levels, and the transmission constraint relationship from planning results to implementation management is not clear enough. The core control content is difficult to implement, which makes the expression ability of planning implementation and management low; Third, the traditional planning control method has a low degree of automation: the current dynamic monitoring and intelligent control have proposed the use of a new generation of artificial intelligence technology to meet the actual needs of intelligent land space planning and control. Faced with the new requirements of dynamic monitoring and intelligent control, how to use a new generation of artificial intelligence technology to meet the actual needs of intelligent land space planning and control and improve the efficiency of planning implementation and management has become a problem that needs to be solved urgently. Summary of the invention
[0003] The purpose of this application is to provide a method, product, medium and equipment for intelligent management and control of national land space planning, which can effectively improve the accuracy, efficiency and visual expression ability of intelligent management and control of national land space planning.
[0004] To achieve the above objectives, this application provides the following solutions.
[0005] In the first aspect, the present application provides a method for intelligent management and control of national land space planning, including: obtaining basic geographic information data and natural resource management data within the target area, and constructing a basic geographic information data sub-database and a natural resource management data sub-database; the basic geographic information data include digital line maps, digital elevation models and digital orthophotos of different spatial scales; the natural resource management data involves vector data of the location, form and attribute information of the spatial control boundary, and survey and statistical data describing various social and economic activities occurring in the national land space.
[0006] Obtain policy and regulatory data within the target area and construct a national land space planning and control knowledge graph sub-library; the policy and regulatory data include relevant management rules in policy documents.
[0007] Deep learning methods are used to perform remote sensing interpretation of the basic geographic information data sub-database and the natural resources management data sub-database, and to construct a thematic data sub-database for remote sensing interpretation of natural resources.
[0008] The latest temporal remote sensing image data within the target area is obtained and input into the natural resources remote sensing interpretation thematic data sub-database to obtain the current land use spatial data.
[0009] Logically judge the current land use spatial data with the corresponding policy and regulation data in the national land space planning and control knowledge graph sub-library to obtain intelligent control results; the intelligent control results include normal and abnormal.
[0010] The three-dimensional visualization technology is used to present the intelligent management and control results in an all-round way in the three-dimensional geographic scene.
[0011] In a second aspect, the present application provides a computer program product, including a computer program, which, when executed by a processor, implements the intelligent management and control method of national land space planning.
[0012] In a third aspect, the present application provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the intelligent management and control method of national land space planning is implemented.
[0013] In a fourth aspect, the present application provides a computer device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the intelligent management and control method for national land space planning.
[0014] According to the specific embodiments provided in this application, this application discloses the following technical effects.
[0015] The present application provides a method, product, medium and equipment for intelligent management and control of national land space planning. By constructing a basic geographic information data sub-base and a natural resource management data sub-base, a spatiotemporal data base with unified spatial benchmarks and standardized data content is formed; then, based on the spatiotemporal data base, a deep learning method is used to perform remote sensing interpretation on it, thereby constructing a natural resource remote sensing interpretation thematic data sub-base; the latest phase remote sensing image data acquired within the target area is input into the natural resource remote sensing interpretation thematic data sub-base, and data is extracted to obtain spatial data on the current land use status, and the data is combined with the constructed national land space planning and control knowledge graph sub-base, compared with policy and legal data, and knowledge is analyzed and mined to form a highly automated and intelligent national land space planning and control smart brain, which effectively improves the accuracy and efficiency of intelligent management and control of national land space planning; finally, three-dimensional visualization technology is used to realize three-dimensional visualization expression of intelligent management and control results. BRIEF DESCRIPTION OF THE DRAWINGS
[0016] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0017] Figure 1 Schematic diagram of the process flow of the intelligent management and control method for land space planning provided for this application.
[0018] Figure 2 Sample entry flow chart provided for this application.
[0019] Figure 3 Intelligent interpretation and processing knowledge graph provided for this application.
[0020] Figure 4 A diagram of the cluster service process provided for this application.
[0021] Figure 5 Schematic diagram of the structure of the intelligent management and control system for national land space planning provided for this application. DETAILED DESCRIPTION
[0022] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.
[0023] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below in conjunction with the accompanying drawings and specific implementation methods.
[0024] The key technical effect that the land space planning intelligent control method disclosed in this application aims to achieve is to form a highly automated and intelligent land space planning control smart brain, and analyze and apply it in a real-life three-dimensional scene. The specific steps are as follows: Figure 1 As shown, the steps include steps 1 to 6.
[0025] Collect basic geographic information data, natural resource management data, policy and regulatory data, etc. according to the target area, process and aggregate various data resources, build a spatiotemporal database for national land space planning, and carry out subsequent operations based on the spatiotemporal database.
[0026] Step 1: Obtain basic geographic information data and natural resource management data within the target area, and build a basic geographic information data sub-database and a natural resource management data sub-database.
[0027] (1) Obtain basic geographic information data and natural resource management data within the target area.
[0028] Obtain basic geographic information data within the target area, including digital line maps, digital elevation models, digital orthophotos, place names and addresses at different spatial scales such as 1:10000, 1:5000, and 1:2000. The main contents of the digital line maps include vector data such as water systems, settlements, transportation, boundaries and political divisions, vegetation, and pipelines.
[0029] Acquire various administrative management data approved, supervised and derived by the natural resources department within the target area, including vector data describing the location, form and attribute information of spatial control boundaries such as ecological protection red lines, permanent basic farmland scope, urban development boundaries, land use approval, sea use approval, mining rights approval, land supply, planning permits, real estate registration, various approved development parks, and survey and statistical data describing various social and economic activities occurring in the national land space.
[0030] (2) Data preprocessing.
[0031] After obtaining the basic geographic information data and natural resource management data within the target area, preprocessing the acquired data can keep the spatial structure of multi-source data such as basic geographic information data and natural resource management data consistent with the geographic reference, providing a reliable data foundation for subsequent analysis and control.
[0032] For survey and statistical data in non-vector data format in the acquired basic geographic information data and natural resource management data within the target area, the preprocessing steps include address standard structuring, search matching, and data format conversion. For other basic geographic information data and natural resource management data in vector data format, the preprocessing steps include unified spatial benchmark, data cleaning and processing, and metadata rule control.
[0033] (3) Data is aggregated and stored in the database.
[0034] Taking into account factors such as data quality, spatiotemporal characteristics, accuracy requirements and computing resources, the pre-processed multi-source data will be uniformly stored and quality controlled to achieve data aggregation and complete the construction of the basic geographic information data sub-database and the natural resources management data sub-database.
[0035] Step 2: Obtain policy and regulatory data within the target area and build a sub-database of the national land space planning and control knowledge graph.
[0036] (1) Extracting land space planning control rules.
[0037] The above-mentioned policy and regulatory data include relevant management rules in policy documents. Based on knowledge graph technology, we capture the laws and regulations related to land and space control and the control rules contained in policy documents, establish a three-level knowledge graph framework of "control policy-control scenario-control element", analyze the control objects and constraints corresponding to the control rules, and form the "entity-relationship-entity" triplet in the professional field of land and space planning.
[0038] (2) Knowledge graph expression of planning and control rules.
[0039] The extracted planning and control rules are structured in the form of a graph data structure, which can not only clearly display the correlation and hierarchical relationship between various entities, but also effectively store and retrieve relevant information of the planning and control rules, thereby completing the construction of the national land space planning and control knowledge graph sub-library.
[0040] Step 3: Use deep learning methods to perform remote sensing interpretation on the basic geographic information data sub-database and the natural resource management data sub-database and construct a natural resource remote sensing interpretation thematic data sub-database.
[0041] (1) Construct a sample set of interpretations of natural resource entities.
[0042] Guided by the control elements in the knowledge graph constructed in step 2, a typical natural resource entity interpretation sample set is constructed through three technical processes: sample design, sample production, and sample storage. Through the standardized storage, indexing, and management of sample data sets, a shared semantic natural resource entity interpretation sample set is formed, laying the foundation for building a natural resource entity pre-training model, establishing an intelligent remote sensing interpretation service model, and realizing the automated and intelligent extraction of control elements.
[0043] ① Sample design, specifically refers to the standardized design of samples to provide guarantees for the intelligent and efficient management and use of samples. The sample categories preset three categories: change detection, object classification, and target detection. The design content includes sample type, type code, type description, etc. Among them, the contents of the three categories of change detection, object classification, and target detection are shown in Table 1, Table 2, and Table 3 respectively.
[0044] Table 1 Change detection sample design table
[0045]
[0046] Table 2 Sample design table for land feature classification
[0047]
[0048] Table 3 Target detection sample design table
[0049]
[0050] ② Sample production, the specific production process includes sample collection and sample production. Before sample collection, it is necessary to collect and organize data related to historical project results to provide a reference data source for subsequent sample collection. Sample collection is to draw the sample vector boundary and mark the information based on the remote sensing image features in the basic geographic information data sub-database and the natural resources management data sub-database, or to screen, refine the map spots and mark the information based on historical data results to form the original sample. Sample production is to make the original sample into a standard sample with unified color depth, unified tone, unified format, and unified spatial reference.
[0051] ③ Sample storage. Due to the diversity of data sources, data types and sample types, there are certain differences in sample directory organization, file naming, expression format, sample category, etc. Therefore, samples need to undergo a series of preprocessing before storage, see Figure 2 , as shown below.
[0052] 1) Data integrity check: Mainly check whether the sample image files to be managed correspond to the label files one by one, and whether the number of sample files is complete.
[0053] 2) Data format inspection: Mainly check whether the data formats of the image files and label files of the samples to be managed meet the storage requirements, and whether the identifiers of the samples to be managed meet the management sample standards.
[0054] 3) Category verification. Check whether the task categories of the samples to be managed (land feature classification, target detection and change detection), the sample land feature type codes and the corresponding identifiers such as spelling are correct. If incorrect, the data needs to be modified and re-checked for data integrity.
[0055] 4) Category mapping: Map the feature type code corresponding to the category in the identifier to the corresponding feature type name to facilitate subsequent sample management, analysis, retrieval and downloading.
[0056] 5) Data format conversion: Convert the sample label shp file to be managed into GeoJson format so that it can be stored in the MongoDB database.
[0057] 6) Sample information acquisition: Read the sample spatial range, resolution, number of bands, sample size, imaging time, annotation time, and sample categories from the sample file identifier to form sample metadata for management and use.
[0058] 7) Sample storage. After completing the above storage operation process, the sample metadata information is organized and imported into the database according to the designed database logic model, and the image and label entity files are imported into the MongoDB database according to the sample organization specifications to construct a natural resource entity class interpretation sample set.
[0059] (2) Construct a pre-trained model for natural resource entities.
[0060] Model training is carried out for typical natural resource entities, and a multi-source evaluation mechanism is established based on the integrity of the entity on the basis of conventional machine learning evaluation indicators. The optimal pre-trained model and metadata information that meet the expected values are stored in the database for management, and a typical pre-trained model set that can be used out of the box is constructed.
[0061] ① Model training. The sample set of natural resource entity interpretation is input into the convolutional neural network model. Through the guidance of geological knowledge, the loss constraint based on geological knowledge reasoning is constructed, and the model training is guided autonomously. The specific steps are as follows.
[0062] 1) Construction of geoscientific knowledge reasoning rules. Using SWRL language to describe geoscientific knowledge reasoning rules can directly correct misclassifications, improve classification accuracy, and satisfy formulas (1) and (2).
[0063] (1).
[0064] (2).
[0065] Where entity represents the correctly classified ontology instance of the reasoning unit, misEntity represents the incorrectly classified ontology instance of the reasoning unit, rs represents the name of the geoscientific knowledge reasoning rule, Semantic Segment represents the attribute of the entity reasoning unit ontology instance, and MisSegmantic Segment represents the attribute of the misEntity reasoning unit ontology instance.
[0066] Geographic knowledge reasoning rules include two categories: spatiotemporal attribute constraints and spatiotemporal relationship constraints. The first category of rules is based on attribute definitions as basic attribute constraints and other explicit constraints; the second category of rules defines the constraint principle of the spatial position and attribute information relationship of the target patch and the adjacent patch, which is applicable to eliminating the hole phenomenon caused by misclassification, that is, correcting the classification of the hole area to the category of the surrounding objects, and is applicable to misclassification caused by inconsistent spatial relationships and shadow information extraction. The specific geographic knowledge reasoning rules are shown in Table 4.
[0067] Table 4 Reasoning rules for geoscience knowledge
[0068]
[0069] 2) Model training guided by geoscientific knowledge. Model training is essentially a data-driven classification method for pixel-wise loss inverse optimization. Through geoscientific knowledge guidance, loss constraints based on geoscientific knowledge reasoning are constructed to autonomously guide model training.
[0070] ②Model performance evaluation. After the model training is completed, the image of the target area needs to be used to evaluate the model performance. The evaluation index adds the image patch connectivity similarity evaluation on the basis of the existing conventional evaluation indicators such as accuracy, precision, recall rate and F1 Score to improve the recognition of remote sensing interpretation quality and keep the performance evaluation index value consistent with the subjective quality score.
[0071] ③ Construction of pre-trained model set. The pre-trained models and metadata information including evaluation indicators are stored in the database for management, a pre-trained model set is built for out-of-the-box use, and the construction of the natural resources remote sensing interpretation thematic data sub-database is completed.
[0072] (3) Establish an intelligent remote sensing interpretation service model.
[0073] After building the natural resources remote sensing interpretation thematic data sub-database, we further established an intelligent remote sensing interpretation service model. The intelligent interpretation service model is based on the core concept of "multi-layer perception, layered application, and on-demand integration", providing data mining product services and resource management and scheduling, and assisting decision-making analysis, including intelligent interpretation processing service model and distributed service cluster service model. Intelligent interpretation processing is used to dynamically analyze image content and provide automated remote sensing knowledge mining capabilities. Distributed service clusters are used to publish data resources and provide data services for intelligent management and control systems.
[0074] The intelligent interpretation and processing service model is an intelligent interpretation service model provided for change detection, target detection and object classification, including three elements: mathematics, physics and images.
[0075] "Data" refers to remote sensing interpretation image data. In the remote sensing interpretation of national land space planning and control, "data" is organized according to project classification and task type interpretation data. Among them, "project classification data" comes from the database, serves the remote sensing interpretation project, and provides classification management for all image data of a certain project; "task type interpretation data" is a collection of data required for the project to interpret a certain task type, which provides an image data source for "task type interpretation"; project classification data and task type interpretation data together constitute project data. "Project classification data" is organized and managed according to region, resolution, time, image source, etc. According to the needs of task type interpretation, view the classification list of all image data of the project from the "project classification data", select and load remote sensing interpretation images that meet the requirements to support the development of interpretation work of different task types.
[0076] "Reasoning" refers to remote sensing interpretation models. In the remote sensing interpretation of national land space planning and control, "reasoning" is manifested as various semantic segmentation models and numerical calculation tools, mainly including DeepLabV3+, PSPNet, Transformer model, raster data cutting algorithm, irregular triangulated network (Triangulated Irregular Network, TIN) and other models and calculation tools. The semantic segmentation model is a model for computer vision tasks that aims to segment an image into different semantic regions, that is, assign each pixel in the image to its corresponding semantic category, help the computer understand each pixel in the image, and provide more accurate information for subsequent analysis and decision-making.
[0077] “Xiang” refers to the image diagram. To achieve the standardization and normalization of intelligent interpretation and processing, a unified image diagram should be established for it, that is, the intelligent interpretation and processing knowledge graph, such as Figure 3As shown in the figure. Intelligent interpretation processing includes three categories: "change detection", "target detection" and "land object classification". Each type of interpretation includes several specific interpretation contents, and the specific interpretation indicators of each content are organized according to "attribute-feature-object". Taking "change detection" interpretation as an example, its interpretation indicator system includes confidence setting, filter area, model selection, etc. The intelligent interpretation processing knowledge graph stipulates the interpretation framework of three task types and establishes a unified indicator system for intelligent interpretation processing services.
[0078] As for the distributed service cluster service model, with the development of sensors, remote sensing platforms, data communications and other related technologies, the amount of data obtained through remote sensing has expanded dramatically. At the same time, space application needs have put forward higher requirements on the speed and accuracy of remote sensing data processing. "Difficult processing" has become an increasingly prominent problem. In order to improve the speed of remote sensing data processing and product accuracy, solve the problem of rapid production of geographic information data, and improve the insufficient computing power of massive remote sensing data in the process of high-performance computing, a cluster processing system is established to provide a distributed service cluster service model.
[0079] The distributed service cluster service mode is used in conjunction with the intelligent interpretation and processing service to form multi-computing node parallel computing capabilities, provide a unified intelligent service address for all operation terminals, and the processing system automatically allocates computing tasks, making full use of the LAN / Internet hardware resource computing power to improve operation production efficiency. The specific operations of cluster services are as follows: Figure 4 shown.
[0080] Step 4: Obtain the latest temporal remote sensing image data within the target area and input it into the natural resources remote sensing interpretation thematic data sub-database to obtain the current land use spatial data.
[0081] After obtaining the latest phase remote sensing image data within the target area, it is input into the above-mentioned intelligent remote sensing interpretation service model established based on the natural resources remote sensing interpretation thematic data sub-database, automatically completing the pixel classification of the image, extracting cultivated land, gardens, woodlands, grasslands, construction land, water areas and other land features to form spatial data on the current land use status.
[0082] Step 5: Perform logical judgment on the current land use spatial data and the corresponding policy and regulatory data in the national land space planning and control knowledge graph sub-library to obtain intelligent control results.
[0083] Since it is necessary to use the target land feature category as an entity keyword to query the relevant control rule knowledge, a comprehensive analysis is conducted by combining the current land use spatial data with the control rule knowledge to preliminarily determine the rationality and compliance of the current land use spatial data. Therefore, it is necessary to provide a query service for planning control rule knowledge.
[0084] The query service is a query of existing rule knowledge, including conditional query and non-conditional query. Conditional query can query entities or relationships with preset conditions. Non-conditional query directly returns all control rule knowledge in a paged form.
[0085] (1) Development of knowledge services for planning and control rules.
[0086] The planning and control rule service adopts the form of application programming interface (API), which is the interface or communication protocol between two different computer programs.
[0087] ① Define API interface parameters. Determine the parameters according to the query service type. If it is a conditional query, define the corresponding query object and query rules; if it is a non-conditional query, the parameters can be left blank.
[0088] ② Organize the query of the graph library. According to the interface parameters, organize the SQL statements and query in the graph library.
[0089] ③Define the return value. Assemble the query results into a structured form and determine the return value.
[0090] (2) Sharing of knowledge services on planning and control rules.
[0091] The developed application program interface is published online for application calls.
[0092] (3) Planning and control rules knowledge service management.
[0093] In order to ensure the security of the service, the authentication and monitoring functions are registered for the knowledge service.
[0094] ① Authentication. In order to solve the problem of interface security, the knowledge service assigns a secret key to each registered application, and the application must carry it every time it calls the API interface.
[0095] ② Monitoring. To ensure data security, monitor the request response time and number of interface calls of the interface, so as to better plan and manage the knowledge service of control rules.
[0096] After obtaining the corresponding policy and regulatory data in the national land space planning and control knowledge graph sub-library through the set query service, the current land use spatial data and the corresponding policy and regulatory data in the national land space planning and control knowledge graph sub-library are logically judged to obtain the intelligent control result; if the current land use spatial data is inconsistent with the corresponding policy and regulatory data in the knowledge graph sub-library, the intelligent control result is abnormal; if the current land use spatial data is consistent with the corresponding policy and regulatory data in the knowledge graph sub-library, the intelligent control result is normal.
[0097] Step 6: Use 3D visualization technology to fully present the intelligent management and control results in a 3D geographic scene.
[0098] The intelligent management and control results extracted in step 5 are deduced in a three-dimensional geographic scene using three-dimensional visualization technology, presenting more intuitive analysis results in all directions, so as to provide intuitive and comprehensive national land space planning and control decision-making support information.
[0099] (1) Construction of three-dimensional scenes for national land space planning.
[0100] The constructed spatiotemporal big database is used as data input, and specific type-driven strategies are formulated for terrain, raster, and vector respectively. The three-dimensional scenes are flexibly assembled as needed to realize the on-demand expression of three-dimensional scenes of national land space planning.
[0101] ① The constructed basic geographic information database sub-database will publish terrain services, image services and vector services according to data types. Terrain services express the basic terrain of three-dimensional scenes, image services express the texture features of three-dimensional scenes, and vector services express other geographic information of three-dimensional scenes on demand.
[0102] ② Establish a data loading optimization strategy for large 3D scenes. Build 3D scenes based on the 3D Earth Engine, consider the smoothness of scene roaming, and adopt the "progressive on-demand loading" optimization strategy to greatly reduce the loading time while taking into account a certain visual quality. The specific steps are as follows.
[0103] 1) 3D data segmentation and simplification: Based on the 3D Tiles standard, the 3D data is spatially segmented and simplified through the multi-levels of detail (LOD) technology.
[0104] 2) Based on the screen space error, select the loaded tile. During the tile loading process, the current Sum of Squares for Error (SSE) value is first calculated, and then compared with the set SSE maximum value. When the actual value exceeds the specified maximum SSE, it indicates that the current level of tile is not fine enough and the next level of sub-tiles needs to be loaded.
[0105] 3) Select optimization options to optimize data loading as needed. Loading optimization options include several main optimization options and sub-items included in some optimization options, as shown in Table 5.
[0106] Table 5 Main optimization options and their descriptions
[0107]
[0108] (2) Three-dimensional analysis and expression of national land space planning and control.
[0109] The constructed natural resource management sub-database will publish vector services according to data types to provide data references at the management level for analysis. The intelligent extraction and mining results of land space remote sensing data obtained through the above steps will be published as vector services and presented in the three-dimensional geographic scene as intelligent management and control results. The three-dimensional analysis expression includes numerical analysis of management and control results and simulation analysis of application scenarios.
[0110] ① Numerical analysis of control results. Use multiple analysis views such as geographic spatial distribution maps, geographic spatial statistical maps, and geographic spatial relationship maps to intuitively present and deeply explore the changes and patterns of control data. For example, use three-dimensional charts such as single column charts, clustered column charts, and stacked column charts to present the quantitative statistical analysis of different land types and the dynamic change trend of the same land type changes over the years; use three-dimensional spatial distribution maps such as node trajectory maps, star maps, and heat maps to present the distribution patterns of monitoring targets.
[0111] ②Simulation analysis of application scenarios. Using information such as three-dimensional terrain, control data, and current construction land distribution, we conduct control simulations of different application scenarios such as rainfall or sunlight, providing reasonable preset scenarios for further deduction and analysis of control results.
[0112] In summary, this application first constructs a large spatiotemporal database of national land space planning to form a spatiotemporal data base with unified spatial benchmarks and standardized data content; secondly, using technologies such as deep learning and knowledge graphs, establish an intelligent remote sensing interpretation service model, extract national land space planning control rules, and build a knowledge service for national land space planning control rules; then, obtain new phase remote sensing image data of the control area, access the intelligent remote sensing interpretation service, extract data, combine the national land space control knowledge graph, compare with the control rule indicators, analyze and mine knowledge, and form a highly automated and intelligent national land space planning control smart brain; finally, through national land space planning analysis and visualization technology, realize the three-dimensional visualization expression of the analysis results. Therefore, this application integrates multiple types of data such as natural resource elements and business rules into a three-dimensional visualization space system, conducts high integration and mining analysis, and constructs intelligent management applications, which improves the visualization analysis, judgment and management planning decision-making capabilities of various monitoring objects in three-dimensional geographic space, and builds a spatiotemporal information technology system with comprehensive dynamic perception, system accurate cognition, and global intelligent control as the main line, which comprehensively improves the national land space governance capabilities.
[0113] This application also provides a specific application scenario, which applies the above-mentioned land space planning intelligent control method. Specifically: Taking the planning intelligent control of the ecological zone around the city of a certain city as an example, the implementation steps of this application are as follows.
[0114] 1. Obtain basic geographic information data of the entire ecological zone around a city, such as the annual land change survey, temporal satellite images, and boundary lines of the ecological zone around the city since 2009, natural resource management data such as cultivated land reserves, permanent basic farmland, urban development boundaries, urban yellow lines, urban blue lines, urban green lines, urban purple lines, earthquake fault zones and other vector data involved in the ecological zone around the city, important control space geological disaster risk zoning, national land space planning zoning, land consolidation and ecological restoration zoning, etc. The spatial mathematical basis of the above data is unified to the 2000 National Geodetic Coordinate System, and the fusion processing forms a basic geographic information sub-database and a natural resource management data sub-database.
[0115] 2. Collect relevant policy documents related to the ecological zone around the city. Policy documents cover the control requirements of various elements of the general national land space, such as farmland protection, comprehensive land improvement and ecological restoration, high-standard farmland construction, intensive and economical use of land, and other relevant management rules specified in policy documents. At the same time, plans are also collected for this special area of the ecological zone around the city. After sorting out the above policy documents, the upper limit constraints on the scale, control space requirements, and core indicator values are extracted in the format of "entity-relationship-entity", the extracted control rules are fused and cleaned, and then the ontology is constructed in the graph database Neo4j, including farmland protection, land use, and boundary control. The control files corresponding to the control rules in the entity node attribute association can realize online file browsing and provide decision support, thereby completing the construction of the national land space planning control knowledge map sub-library.
[0116] 3. According to the actual requirements of the management and control of the ecological zone around the city, the preset natural resource remote sensing interpretation thematic database can support the automatic classification and extraction of objects in a single period of remote sensing images. Therefore, a natural resource entity sample data set including object classification samples and a natural resource entity pre-trained model data set including object classification pre-trained models are built. The sample data set comes from historical project results data, and the sample standardization format is transformed. The pre-trained model set is a pre-trained model with typical sample feature learning ability obtained by inputting the built sample data set into the convolutional neural network through long-term iterative learning and training, thereby completing the construction of the natural resource remote sensing interpretation thematic data sub-library.
[0117] 4. Serve the pre-trained model with typical sample feature learning capabilities, build a land feature classification processing service, and then deploy the land feature classification processing service to multiple computing nodes to form a distributed service cluster interpretation service.
[0118] 5. Develop an application programming interface based on the national land space planning and control knowledge graph library, publish it online for subsequent application scenarios to call, and provide query services for planning and control rule knowledge.
[0119] Based on the knowledge graph library of the ecological zone around the city, on the one hand, it is possible to directly query the relevant management and control rules and source files, quickly compare the different management and control requirements of various levels of planning, and improve the scientific accuracy of management; on the other hand, the rules extracted from the knowledge graph can be used to clarify the spatial access requirements within the ecological zone around the city. For example, whether urban construction projects can access the farmland protection zone planned by the ecological zone around the city, the project scope can be imported into the platform, and the access scenario can be selected so that the platform can preliminarily judge the access conditions as a tool to assist decision-making.
[0120] 6. After completing the construction of the spatiotemporal big database of national space planning and the establishment of remote sensing intelligent interpretation and planning control services, obtain the 2023 high-current remote sensing image data of the surrounding ecological zone, establish the permanent basic farmland control task, and monitor and warn the permanent basic farmland spots according to the control requirements entered in the knowledge map. If the spatial access requirements of the permanent basic farmland spots are met, but the image interpretation finds that the amount of cultivated land in the permanent basic farmland has decreased, and rural roads, ditches and other land types have appeared. The permanent basic farmland in the high-tech South District of the surrounding ecological zone involves 20 mu of rural roads and ditches, and the current status is non-cultivated land. Feedback the identification results to the municipal functional management department in a timely manner. After research by the competent department, the area is implementing high-standard farmland construction, which meets the permanent basic farmland management regulations, but the project approval and vector range have not yet been included in the scope of the big database of this platform. At the same time, the output results warn of the planting of crops in some spots, and there is a suspected non-planting access requirement, which has been fed back to the agricultural competent department in a timely manner.
[0121] 7. Use three-dimensional visualization technology to deduce the intelligent control results in a three-dimensional geographic scene, presenting more intuitive analysis results to provide scientific land space planning and control decision-making support information.
[0122] In response to the actual needs of national land space planning status assessment and analysis, planning implementation supervision and intelligent management and control, this application uses knowledge graphs, three-dimensional analysis and intelligent reasoning to strengthen the effective transmission between various levels of planning and between planning and implementation, forming an information closed loop of "planning-implementation-status", and realizing the intelligent management and control method of "three-dimensional expression, dynamic analysis, and intelligent management and control", and comprehensively promoting the transformation and upgrading of national land space planning and control. Therefore, compared with traditional national land space planning and control technologies that usually rely on manual processing and analysis based on statistical methods, this application has the following significant advantages.
[0123] (1) Data comprehensiveness and consistency.
[0124] Traditional methods usually rely on limited field surveys, statistics and manual data collation, which are easily limited by the problems of incomplete and inconsistent data. The method of this application uses multi-source heterogeneous spatiotemporal big data to build a geospatial database, forming a unified spatiotemporal data base, ensuring the comprehensiveness and consistency of the data.
[0125] (2) Application of high-resolution remote sensing data.
[0126] Traditional methods often use low-resolution maps or hand-drawn drawings, which limit the capture of details. The method of this application uses high-resolution remote sensing data to obtain the latest phase of images, improves the quality and spatial resolution of the data, and makes planning more accurate and detailed.
[0127] (3) Intelligent interpretation and knowledge graph.
[0128] Traditional methods rely on professionals to manually interpret maps or images, which is time-consuming and laborious. This application method combines deep learning with knowledge graphs to design an intelligent remote sensing interpretation service model to improve the intelligent analysis and interpretation capabilities of remote sensing data.
[0129] (4) Automated processing and intelligent brain.
[0130] Traditional methods usually require manual intervention and are inefficient. The method in this application uses deep learning methods to extract and mine massive land space remote sensing data, forming a highly automated and intelligent land space planning and control brain, which improves processing efficiency.
[0131] (5) Three-dimensional planning analysis and visualization.
[0132] Traditional methods often use two-dimensional maps, which are difficult to express planning information intuitively. The present application method uses three-dimensional geographic information planning analysis and visualization technology to make planning results more intuitive and easy to understand.
[0133] In general, the intelligent management and control method of national land space planning in this application emphasizes data comprehensiveness, intelligent interpretation and analysis, automated processing and three-dimensional visualization through the comprehensive use of multiple advanced technologies. Compared with traditional methods, it has higher accuracy, efficiency and visual expression capabilities.
[0134] Specifically, based on the method described in this application, Figure 5 As shown, the present application also discloses an intelligent management and control system for national land space planning, which includes a natural resource intelligent interpretation unit, a planning knowledge graph unit and a three-dimensional visualization unit. Through the establishment of an intelligent management and control system, the intelligent management and control effect of national land space planning of "three-dimensional expression, dynamic analysis, and intelligent management and control" can be achieved.
[0135] Among them, the natural resource intelligent interpretation unit uses image data as input and interpretation vector results as output, and is used to dynamically analyze image content and provide automated remote sensing knowledge mining capabilities. The planning and control knowledge graph unit is used for associated queries or compound queries of planning and control rules, and provides planning and control knowledge intelligent analysis and search capabilities. The three-dimensional visualization analysis unit is used to present planning and control results in three-dimensional geographic scenes, and restore planning and control scenes from a real-world perspective.
[0136] In addition, the present application also provides a computer program product, including a computer program, which, when executed by a processor, implements the steps of the intelligent management and control method for national land space planning.
[0137] Furthermore, the present application also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the intelligent management and control method of national land space planning are implemented.
[0138] Furthermore, the present application also provides a computer device, including a processor, a memory, an input / output interface (I / O for short), a communication interface, and a computer program stored in the memory and executable on the processor, and the processor executes the computer program to implement the steps of the intelligent management and control method of national land space planning.
[0139] Specifically, a land space planning intelligent management and control device includes a data collection end, a server end and a client end. Among them, the server end is composed of intelligent interpretation and processing services, planning and control knowledge services, and distributed service clusters. The intelligent interpretation and processing service is used to dynamically analyze image content and provide automated remote sensing knowledge mining capabilities; the distributed service cluster is used to publish data resources and provide data services for the intelligent management and control system; the client end is composed of several personal computers equipped with independent graphics processing units (GPUs).
[0140] Wherein, the processor, memory and input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Wherein, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store pending transactions. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, the intelligent management and control method of land space planning is implemented.
[0141] It should be noted that the object information (including but not limited to object device information, object personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the object or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant laws, regulations and standards of relevant countries and regions.
[0142] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited to this.
[0143] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.
[0144] This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and core ideas of this application. At the same time, for those skilled in the art, according to the ideas of this application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.
Claims
1. A method for intelligent management and control of national land space planning, characterized in that: include: Obtain basic geographic information data and natural resource management data within the target area, and build a basic geographic information data sub-database and a natural resource management data sub-database; The basic geographic information data includes digital line maps, digital elevation models and digital orthophotos of different spatial scales; The natural resource management data mentioned above involves vector data on the location, form and attribute information of the spatial control boundary, as well as survey and statistical data describing various social and economic activities occurring in the national territory; Obtain policy and regulatory data within the target area and construct a knowledge graph sub-library of land space planning and control; the policy and regulatory data include relevant management rules in policy documents; The deep learning method is used to perform remote sensing interpretation on the basic geographic information data sub-database and the natural resource management data sub-database, and to construct a natural resource remote sensing interpretation thematic data sub-database, including: Obtain original samples based on the basic geographic information database sub-database and the natural resources management database sub-database; The original samples are processed with unified color depth, unified tone, unified format and unified space reference to obtain standard samples; After preprocessing, the standard samples are imported into the MongoDB database to construct a natural resource entity interpretation sample set, which includes: After the standard samples are processed in sequence, including data integrity check, data format check, category check, category mapping, data format conversion, and sample information acquisition, they are imported into the MongoDB database according to the sample organization specifications to construct a natural resource entity class interpretation sample set; The sample set of natural resource entity interpretation is input into the convolutional neural network model for remote sensing interpretation, and the construction of the natural resource remote sensing interpretation thematic data sub-database is completed; Obtain the latest temporal remote sensing image data within the target area and input it into the natural resources remote sensing interpretation thematic data sub-database to obtain the current land use spatial data; Perform logical judgment on the spatial data of current land use status and the corresponding policy and regulation data in the knowledge graph sub-library of national land space planning and control to obtain intelligent control results; the intelligent control results include normal and abnormal; The three-dimensional visualization technology is used to present the intelligent management and control results in an all-round way in the three-dimensional geographic scene.
2. The intelligent management and control method for national land space planning according to claim 1 is characterized in that: The step of obtaining basic geographic information data and natural resource management data within the target area and constructing a basic geographic information data sub-library and a natural resource management data sub-library specifically includes: Perform address standard structuring, search and matching, and data format conversion preprocessing on the non-vector data in the acquired basic geographic information data and natural resource management data within the target area, and perform unified spatial benchmarking, data cleaning and processing, and metadata rule control preprocessing on the vector data; The preprocessed data will be uniformly stored and quality controlled to complete the construction of the basic geographic information data sub-database and the natural resources management data sub-database.
3. The intelligent management and control method for national land space planning according to claim 1 is characterized in that: The obtaining of policy and regulatory data within the target area and the construction of a land space planning and control knowledge graph sub-library specifically include: Based on the policy and regulatory data obtained within the target area, knowledge graph technology is used to capture the planning and control rules contained in the land and space control policy documents, and the extracted planning and control rules are structured in the form of a graph data structure to store and retrieve relevant information of the planning and control rules, thus completing the construction of the land and space planning and control knowledge graph sub-library.
4. The intelligent management and control method for national land space planning according to claim 1 is characterized in that: The above-mentioned logical judgment of the spatial data of the current land use situation and the corresponding policy and regulation data in the knowledge graph sub-library of national land space planning and control is performed to obtain the intelligent control results, which specifically include: Use the application programming interface (API) to query the knowledge graph sub-database and obtain the corresponding policy and regulatory data; Logical judgment is performed on the current land use spatial data and the corresponding policy and regulatory data in the national land space planning and control knowledge graph sub-library to obtain intelligent control results; if the current land use spatial data and the corresponding policy and regulatory data in the national land space planning and control knowledge graph sub-library are inconsistent, the intelligent control result is abnormal; if the current land use spatial data and the corresponding policy and regulatory data in the national land space planning and control knowledge graph sub-library are consistent, the intelligent control result is normal.
5. The intelligent management and control method for national land space planning according to claim 1 is characterized in that: The use of three-dimensional visualization technology to comprehensively present the intelligent management and control results in a three-dimensional geographic scene specifically includes: Construct a three-dimensional geographic scene for national land space planning based on the basic geographic information database sub-database and the natural resources management database sub-database; The three-dimensional visualization technology is used to present the intelligent management and control results in an all-round way in the three-dimensional geographic scene.
6. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, it implements the intelligent management and control method for national land space planning described in any one of claims 1 to 5.
7. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, it implements the intelligent management and control method for national land space planning described in any one of claims 1 to 5.
8. A computer device comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the intelligent management and control method for land space planning as described in any one of claims 1 to 5.
