Methods, devices, equipment, media, and products for identifying and retroactively recognizing historically lost arable land.

Abstract

This application discloses a method, apparatus, equipment, medium, and product for identifying and retroactively recognizing historical outflow of cultivated land, belonging to the field of land and resources survey and monitoring technology. The method includes: constructing a candidate base map of historical outflow of cultivated land based on multi-time-series land change survey vector data; generating a constraint layer based on multiple spatial constraint data; performing spatial intersection elimination based on the candidate base map of historical outflow of cultivated land and the constraint layer to obtain a candidate set after avoidance; judging the integrity of the topsoil based on the candidate set after avoidance and soil survey data to obtain a set of intact topsoil patches; identifying crop types in the set of intact topsoil patches based on multispectral remote sensing data to obtain a crop type identification result set; performing concentrated contiguous analysis on the patches in the crop type identification result set to obtain a contiguous optimization result set; and performing post-processing on the contiguous optimization result set to generate a cultivated land retroactive vector layer. The embodiments of this application can effectively improve the efficiency and accuracy of identifying and retroactively recognizing historical outflow of cultivated land.

Description

Technical Field

[0001] This application relates to the field of land and resources survey and monitoring technology, and in particular to a method, device, equipment, medium and product for identifying and retroactively recognizing historical outflow of cultivated land. Background Technology

[0002] Arable land is the cornerstone of national food security, and accurately grasping the baseline of arable land resources is a prerequisite for achieving refined protection and management of arable land. Historical outflow of arable land mainly includes the following two situations: First, arable land that was actually occupied for planting trees, fruit trees, tea, etc., before the "Third National Land Survey," but did not meet the standards for forest land or orchard land at the time of the "Third National Land Survey" and was surveyed as arable land. After the "Third National Land Survey," it gradually grew into forest and met the standards for forest land or orchard land, and was surveyed as forest land or orchard land, resulting in the outflow of arable land. Second, arable land outflow caused by agricultural restructuring. According to Document No. 128, historically outflowed arable land that meets the new arable land survey rules can be given priority for retroactive recognition as arable land, surveyed as arable land in the following year, and strictly protected. This policy provides an institutional basis for the retroactive recognition of existing arable land resources.

[0003] Currently, existing technologies lack identification solutions for retrospectively recognizing historically lost cultivated land. Related identification work still relies on manual on-site verification and subjective judgment, which is inefficient and makes it difficult to ensure the consistency of results. Furthermore, the temporal evolution characteristics and spatial correlation attributes of vector data such as land use change surveys are not fully explored, failing to transform historical land use information into effective support for specialized identification. Summary of the Invention

[0004] The purpose of this application is to provide a method, apparatus, equipment, medium, and product for identifying and tracing historical outflow of arable land, which can effectively improve the efficiency and accuracy of identifying and tracing historical outflow of arable land.

[0005] To achieve the above objectives, a first aspect of this application provides a method for identifying and retroactively recognizing historically outflowed arable land, comprising: Based on multi-time series land change survey vector data, a candidate base map of historical outflow of cultivated land is constructed; A constraint layer is generated based on multiple spatial constraint data. Spatial intersection elimination is performed based on the historical outflow candidate base map and the constraint layer to obtain the candidate set after avoidance. Based on the candidate set after avoidance and the soil survey data, the integrity of the topsoil is judged to obtain a set of intact topsoil patches; Crop type identification is performed on the intact patch set of the topsoil layer based on multispectral remote sensing data to obtain a crop type identification result set; A concentrated contiguousness analysis was performed on the patches in the crop type identification result set to obtain a contiguous optimization result set; The contiguous optimization result set is subjected to patch post-processing to generate a cultivated land retrospective vector layer.

[0006] Compared with existing technologies, the method for identifying and retroactively recognizing historically outflowed arable land provided in this application has the following advantages: It constructs candidate base maps of historically outflowed arable land using multi-temporal land change survey vector data, enabling accurate tracing of historical land use information for relevant plots; it conducts spatial intersection elimination based on multiple spatial constraint data, effectively avoiding conflict areas related to planning and ecological management; it combines soil survey data to determine the integrity of the topsoil, ensuring that selected plots have the potential for arable land restoration; it uses multispectral remote sensing data to identify crop types, improving the accuracy of crop type determination; it further optimizes the layout of plots through centralized contiguous analysis to adapt to agricultural production needs; and finally, it generates a arable land retroactive vector layer through plot post-processing, achieving batch and accurate retroactive recognition of historically outflowed arable land. This significantly improves the efficiency and standardization of arable land retroactive recognition work, providing reliable technical support for the refined management of arable land resources.

[0007] In some embodiments, constructing a candidate base map of historical outflow of cultivated land based on multi-time series land change survey vector data includes: Obtain vector data from land use change surveys for multiple consecutive years; Select map patches whose land type was cultivated land in the first base year; From the map patches, those whose land use type changed to orchard or woodland in any subsequent monitoring year are selected and output as candidate base maps of historical outflow of cultivated land.

[0008] In some embodiments, the determination of the integrity of the topsoil layer requires that the following conditions be met simultaneously: The soil texture is within the preset suitable tillage range; The effective soil layer thickness shall not be less than the preset soil layer thickness threshold. The soil organic matter content shall not be lower than the lower limit of the background value of soil organic matter in the pre-defined area; The soil pH value is within the preset range suitable for crop growth.

[0009] In some embodiments, the step of identifying crop types from the intact topsoil patches based on multispectral remote sensing data to obtain a crop type identification result set includes: Construct a remote sensing identification sample reference set for shallow-rooted crops, wherein the shallow-rooted crops include at least shallow-rooted fruits and other shallow-rooted crop categories; Acquire multi-temporal multispectral remote sensing images of the region corresponding to the intact topsoil patch set and perform preprocessing; Extract vegetation feature parameters from the preprocessed remote sensing images; The extracted feature parameters are compared with the remote sensing identification sample reference set for similarity, and the crop type is determined according to the confidence threshold to obtain the crop type identification result set.

[0010] In some embodiments, the step of performing a concentrated contiguous analysis on the patches in the crop type identification result set to obtain a contiguous optimization result set includes: For each patch in the crop type identification result set, the following judgments are performed sequentially: If the area of ​​the patch is greater than the first area threshold, it is determined to be a patch that meets the contiguous condition. Otherwise, if the combined area of ​​the patch and existing cultivated land and permanent basic farmland patches within its preset range is greater than the second area threshold, it is determined to be a patch that meets the contiguous condition. All the image patches that are determined to meet the contiguous area conditions are summarized to form the contiguous area optimization result set.

[0011] In some embodiments, the step of performing patch post-processing on the contiguous optimization result set to generate a cultivated land retrospective vector layer includes: Based on the contiguous optimization result set and the preset minimum area threshold, small patches with an area smaller than the preset minimum area threshold are removed to construct a clean patch set. Based on the cleaned patch set and preset geometric correction rules, the self-intersection topological error of the patch boundary is corrected, and the patch that meets the narrow and long judgment condition is fused or clipped to construct the geometrically corrected patch set. Based on the geometrically corrected map patch set, crop type identification results, topsoil integrity judgment conclusions, and multi-source auxiliary verification data, the disposal method and corresponding disposal reasons for each map patch are determined. Combining the historical land type information, ownership information, area data, topsoil status, crop type, disposal method, and disposal reasons for each map patch, a cultivated land retroactive vector layer is constructed and output.

[0012] To achieve the above objectives, a second aspect of this application provides a device for identifying and tracing historical outflow of cultivated land, the device comprising: The module is used to construct candidate base maps of historical outflow of cultivated land based on multi-time series land change survey vector data; The elimination module is used to generate a constraint layer based on multiple spatial constraint data, and to perform spatial intersection elimination based on the historical outflow candidate base map and the constraint layer to obtain a candidate set after avoidance. The judgment module is used to judge the integrity of the topsoil based on the candidate set after avoidance and the soil survey data, and obtain the topsoil integrity patch set; The identification module is used to identify crop types from the intact patch set of the topsoil layer based on multispectral remote sensing data, and obtain a crop type identification result set. The analysis module is used to perform a concentrated contiguous analysis on the patches in the crop type identification result set to obtain a contiguous optimization result set; The generation module is used to perform post-processing of the contiguous optimization result set to generate a cultivated land retrospective vector layer.

[0013] To achieve the above objectives, a third aspect of this application provides an electronic device, the electronic device including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor executes the computer program to implement the method described in the first aspect.

[0014] To achieve the above objectives, a fourth aspect of the present application provides a computer-readable storage medium comprising a stored computer program, wherein the computer program, when executed, controls the device containing the computer-readable storage medium to perform the method described in the first aspect.

[0015] To achieve the above objectives, a fifth aspect of the present application provides a computer program product, which includes a computer program or computer instructions, wherein the computer program or computer instructions, when executed by a processor, implement the method described in the first aspect. Attached Figure Description

[0016] Figure 1 This is a flowchart of a method for identifying and retroactively recognizing historical outflow of arable land provided in an embodiment of this application; Figure 2 This is a schematic diagram of a historical outflow farmland identification and tracking device provided in an embodiment of this application; Figure 3 This is a schematic diagram of the hardware structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0017] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0018] Arable land is the cornerstone of national food security, and accurately grasping the baseline of arable land resources is a prerequisite for achieving refined protection and management of arable land. In recent years, some regions have experienced the outflow of arable land due to its conversion to non-grain or non-agricultural uses, forming historical outflows of arable land. According to the "Notice on Optimizing the Application of Annual Land Change Survey Results" (Natural Resources Development

[2025] No. 128) jointly issued by the Ministry of Natural Resources and the National Forestry and Grassland Administration, historical outflows of arable land mainly include the following two situations: First, arable land that was actually occupied for planting trees, fruits, tea, etc. before the "Third National Land Survey," but had not yet met the standards for forest land or orchard at the time of the "Third National Land Survey" and was surveyed as arable land, but gradually grew into forests after the "Third National Land Survey" and met the standards for forest land or orchard, and was surveyed as forest land or orchard, resulting in the outflow of arable land; Second, arable land outflow caused by agricultural restructuring. According to the requirements of Document No. 128, historical outflows of arable land that meet the new arable land survey rules can be given priority to be re-identified as arable land, surveyed as arable land in the following year, and strictly protected. This policy provides an institutional basis for the retroactive recognition of existing arable land resources.

[0019] Currently, the technology field related to arable land resource management has formed a certain application foundation, mainly concentrated in two major directions: First, in terms of arable land identification and dynamic monitoring, existing technologies mostly use multi-temporal multispectral remote sensing images as the main data source, combined with image processing algorithms (such as radiometric calibration, atmospheric correction, and image fusion) and machine learning models (such as classification and recognition algorithms) to achieve automated interpretation of land types. The focus is on the accurate extraction of the current arable land area, tracking of land type change trajectories, and monitoring of the current status, providing basic data support for the dynamic supervision of arable land resources. Second, in terms of arable land resource management and control, some technologies integrate multi-source information such as land change survey vector data, soil monitoring data, and permanent basic farmland data to carry out arable land quantity statistics, quality grade assessment, and contiguous analysis, providing technical support for the implementation of arable land occupation and compensation balance, large-scale farming planning, and arable land quality protection.

[0020] However, existing technologies still have significant shortcomings in addressing the specific technical needs of identifying and retroactively recognizing historically lost arable land: First, there is a lack of a dedicated identification and retroactive recognition technology system. Existing methods mostly focus on monitoring current arable land or analyzing general land use changes, making it difficult to adapt to the entire process of tracing historical land use, matching policy conditions, and verifying compliance. This results in current retroactive recognition work still heavily relying on traditional methods such as manual on-site verification, which is not only inefficient but also suffers from inconsistent judgment standards, poor result consistency, and difficulty in large-scale application. Second, there is insufficient targeted integration and application of multi-source data. The temporal evolution characteristics (such as changes in land use codes over the years) and spatial correlation attributes of land use change survey vector data have not been fully explored, and a precise matching mechanism between soil survey data and policy requirements has not been formed, resulting in the data value not being effectively transformed into technical support for retroactive recognition. Third, there is a lack of systematic connection between technical logic and spatial planning and control requirements. Existing technologies have not fully incorporated rigid constraints such as urban development boundaries, ecologically sensitive areas, and geological disaster risk areas, making it easy for technical results to conflict with land and space planning, thus limiting their adaptability and practical application value.

[0021] Based on this, embodiments of this application provide a method, apparatus, equipment, medium, and product for identifying and tracing historical outflow of arable land, which can effectively improve the efficiency and accuracy of identifying and tracing historical outflow of arable land.

[0022] Please see Figure 1 , Figure 1 This is an optional flowchart of the historical outflow farmland identification and retroactive recognition method provided in the embodiments of this application. Figure 1 The method may include, but is not limited to, steps S101 to S106.

[0023] Step S101: Construct a candidate base map of historical outflow of cultivated land based on multi-time series land change survey vector data; Step S102: Generate a constraint layer based on multiple spatial constraint data, and perform spatial intersection elimination based on the historical outflow candidate base map and constraint layer to obtain the candidate set after avoidance. Step S103: Based on the candidate set after avoidance and the soil survey data, the integrity of the topsoil is judged to obtain the topsoil integrity patch set; Step S104: Based on multispectral remote sensing data, crop type identification is performed on the intact topsoil patches to obtain a crop type identification result set; Step S105: Perform a concentrated contiguous analysis on the patches in the crop type identification result set to obtain a contiguous optimization result set; Step S106: Perform patch post-processing on the contiguous optimization result set to generate a cultivated land retrospective vector layer.

[0024] Steps S101 to S106 of this application embodiment involve constructing a candidate base map of historically outflowing arable land using multi-time-series land change survey vector data. This allows for precise tracing of historical land use information for relevant map patches. By relying on multiple spatial constraint data to perform spatial intersection elimination, conflict areas related to planning and ecological management can be effectively avoided. Combining soil survey data to determine the integrity of the topsoil layer ensures that the selected map patches have the potential for arable land restoration. Crop type identification is completed using multispectral remote sensing data, improving the accuracy of crop type determination. Then, the map patch layout is optimized through centralized contiguous analysis to adapt to agricultural production needs. Finally, a arable land retrospective vector layer is generated through map patch post-processing, enabling batch and accurate retrospective identification of historically outflowing arable land. This significantly improves the efficiency and standardization of arable land retrospective work and provides reliable technical support for the refined management of arable land resources.

[0025] In step S101 of some embodiments, multi-time series land change survey vector data refers to land change survey vector data for multiple consecutive years (e.g., 2019-2024). Each map patch contains key attribute fields such as land type code and ownership information, which can trace the historical evolution of land types. The historical outflow candidate base map refers to the set of map patches that meet the condition of "the base year was cultivated land and subsequent years were converted to non-cultivated land (orchard / forest land)", which is the basic dataset for subsequent screening.

[0026] In some embodiments, a candidate base map of historically outflowing cultivated land is constructed based on multi-time-series land change survey vector data, including: Obtain vector data from land use change surveys for multiple consecutive years; Select map patches whose land type was cultivated land in the first base year; From the map patches, select those whose land use type changed to orchard or woodland in any subsequent monitoring year, and output them as candidate base maps of historical outflow of cultivated land.

[0027] In other words, by acquiring multi-temporal land change survey vector data from consecutive years, and through data preprocessing, temporal overlay, and logical querying, map features that were "cultivated land in the baseline year and subsequently converted to orchards / forest land" are selected to construct a candidate base map of historical cultivated land outflow, thus achieving preliminary and accurate tracing of historical cultivated land outflow. Specifically, firstly, multi-temporal land change survey vector data from consecutive years are acquired, a unified coordinate system is established, and the data is cropped to the target area; then, map features with land type codes of cultivated land (e.g., code "01") in the first baseline year (e.g., 2019) are selected. Further, this type of map feature is traversed, and through spatial overlay and temporal correlation, map features with land type codes converted to orchards (e.g., code "02") or forest land (e.g., code "03") in any subsequent monitoring year (e.g., 2020-2024) are selected, ultimately forming a candidate base map of historical cultivated land outflow.

[0028] The specific implementation process is as follows: First, acquire land use change survey vector data (i.e., change data) for multiple consecutive years. This data needs to cover the target area from the first base year to subsequent monitoring years (e.g., six consecutive years from 2019 to 2024), and each map patch needs to contain key attribute fields such as land use code. Simultaneously, acquire reference data such as permanent basic farmland, urban development boundaries, industrial block lines, ecologically sensitive areas, and geological disaster risk areas to assist in data filtering. Then, preprocess all acquired data. First, unify the coordinate system, transform all data to the same projection coordinates, and determine the same elevation datum, projection method, and degree zone to ensure spatial analysis accuracy. Then, use the administrative division boundary vector file of the target area to trim out the change survey data within the required range. Next, perform temporal overlay and logical query. Spatially align and temporally correlate the change data for multiple consecutive years, construct a spatial database, and then execute a logical query: first, filter out the map patch set whose land use code is cultivated land (corresponding to land use code "01") in the first base year (e.g., 2019). Then traverse the image set. For each map patch, spatial overlay analysis is used to query its land use code in any subsequent monitoring year (e.g., 2020-2024). Finally, map patches that have changed their land use attribute to orchard (land use code "02") or forest (land use code "03") in at least one of the subsequent monitoring years are selected. These map patches that meet the time-series change logic are output, thus obtaining the candidate base map of historical outflow of cultivated land. .

[0029] In step S102 of some embodiments, the multiple spatial constraint data refers to the rigid control data in land spatial planning, including data on urban development boundaries, industrial block lines, ecologically sensitive areas, and geological disaster risk areas, which are the main basis for ensuring that the retroactive recognition of cultivated land meets the spatial control requirements. The constraint layer refers to the standardized spatial layer formed after preprocessing the multiple spatial constraint data such as coordinate system 1 and range clipping. The candidate set after avoidance refers to the set of candidate map patches that meet the spatial planning requirements after removing map patches that have spatial conflicts with the constraint layer.

[0030] Constraint layers are generated based on multiple spatial constraints such as urban development boundaries and ecologically sensitive areas. By spatially intersecting and eliminating candidate base maps of historical outflow of cultivated land with the constraint layers, map patches that do not meet spatial control requirements are removed, resulting in a candidate set after avoidance, thus achieving a rigid connection between technical solutions and spatial planning.

[0031] In some embodiments, the specific implementation process is as follows: First, the multiple spatial constraint data specifically includes constrained planning and risk spatial data such as urban development boundaries, industrial block lines, ecologically sensitive areas, and geological disaster risk areas. By acquiring this type of data and performing preprocessing operations such as coordinate unification and range regularization, a standardized constraint layer is generated, providing a unified benchmark for subsequent spatial conflict detection. Then, the constructed historical outflow candidate base map is... Spatial intersection analysis was performed on each of the above constraint layers one by one. The specific operation method was as follows: the candidate base map of historical outflow of cultivated land was used. Spatial extent comparison is performed with any constraint layer. If a patch in the candidate base map intersects with the spatial extent of a constraint layer, it is determined that the patch violates the planning or ecological baseline and is removed from the historical outflow of cultivated land candidate base maps. Finally, all patches that do not spatially conflict with any constraint layer are compiled to form a post-avoidance candidate set. This ensures that all subsequently selected map features meet the rigid requirements of spatial planning and ecological protection.

[0032] In step S103 of some embodiments, the soil survey data refers to the vector data from the Third National Soil Survey, which includes attribute indicators such as soil texture, effective soil layer thickness, soil organic matter content, soil bulk density, and soil pH value. This data is crucial for determining the state of the topsoil. The topsoil integrity assessment refers to a comprehensive judgment of the soil attributes of a plot based on preset standards to confirm whether it meets the conditions for restoration to arable land. The set of plots with intact topsoil refers to the set of plots that meet preset conditions after the topsoil integrity assessment.

[0033] It should be noted that the following conditions must be met simultaneously to determine the integrity of the topsoil: The soil texture is within the preset suitable tillage range; The effective soil layer thickness shall not be less than the preset soil layer thickness threshold. The soil organic matter content shall not be lower than the lower limit of the background value of soil organic matter in the pre-defined area; The soil pH value is within the preset range suitable for crop growth.

[0034] Furthermore, firstly, select the map patches marked with the "recoverable" attribute from the candidate set after avoidance; then, combine the soil survey data to extract the soil attribute indicators of such map patches, and make a comprehensive judgment according to preset conditions (soil texture is within the suitable cultivation range, effective soil layer thickness is ≥ preset threshold, soil organic matter content is not lower than the lower limit of the regional background value, and soil pH value is within the suitable growth range of crops) to select map patches with intact cultivated layer.

[0035] In one specific embodiment, firstly, from the post-avoidance candidate set Map patches marked as "recoverable" in 2024 were selected. This attribute indicates the potential of these plots to be restored to arable land without complex engineering measures, providing a basic screening basis for assessing the integrity of the topsoil. Subsequently, vector data from the Third National Soil Census were obtained, and corresponding areas were cropped according to the selected map patch ranges. Soil property indicators such as soil texture, effective soil layer thickness, soil organic matter content, and soil pH were extracted as key data support for assessing the integrity of the topsoil. The assessment of topsoil integrity requires the following preset conditions to be met simultaneously: the soil texture is within the preset suitable arable texture range, i.e., it has not changed to extreme types such as sandy soil and still maintains a texture suitable for agricultural cultivation; the effective soil layer thickness is not less than the preset soil layer thickness threshold (e.g., 10 cm, which can be slightly adjusted according to regional standards); the soil organic matter content is not lower than the preset lower limit of the background value of arable soil organic matter in the region, and has not experienced a precipitous decline due to human activities; the soil pH is within the preset suitable pH range for crop growth (e.g., 6.0-8.5), and there has been no severe acidification or alkalization. By analyzing the candidate set after the avoidance After preliminary screening and multi-dimensional verification of soil survey data, all plots that meet the above-mentioned criteria for topsoil integrity are compiled to obtain a set of plots with intact topsoil. .

[0036] In step S104 of some embodiments, multispectral remote sensing data refers to satellite remote sensing images (such as Sentinel-2 images) that cover the key growth stages of crops, have a spatial resolution of not less than 10 meters, and include multispectral bands such as blue, green, red, and near-infrared, which can capture the spectral and phenological characteristics of crops. Crop type identification refers to the process of determining the crop type by extracting vegetation feature parameters from the map patches and comparing them with a preset sample reference set for similarity. The crop type identification result set refers to a complete dataset that includes automatically determined results (such as specific shallow-rooted crop types), low-confidence results awaiting manual determination, and results of greenhouse crops awaiting manual determination.

[0037] Based on multispectral remote sensing data, crop type identification was carried out on the intact topsoil patch set by constructing a shallow-root crop sample reference set, preprocessing images, extracting vegetation feature parameters and comparing similarity. The result set of crop type identification results, which includes automatic judgment results and results to be manually determined, was obtained, which accurately matched the crop type requirements for farmland retroactive identification.

[0038] In some embodiments, crop type identification is performed on a set of intact topsoil patches based on multispectral remote sensing data to obtain a crop type identification result set, including: Construct a remote sensing identification sample reference set for shallow-rooted crops, which includes at least shallow-rooted fruits and other shallow-rooted crop categories; Acquire multi-temporal multispectral remote sensing images of the regions corresponding to the intact topsoil patch set and perform preprocessing; Extract vegetation feature parameters from the preprocessed remote sensing images; The extracted feature parameters are compared with the remote sensing identification sample reference set for similarity, and the crop type is determined based on the confidence threshold to obtain the crop type identification result set.

[0039] Specifically, firstly, a remote sensing identification sample reference set for shallow-rooted crops is constructed. This reference set can be determined by referring to relevant technical specifications for farmland use, crop identification standards, or management regulations corresponding to the target area. The target crop range includes four main categories: medicinal herbs, shallow-rooted fruits, shallow-rooted flowers, and other shallow-rooted crops. A detailed crop list is shown in Table 1. Among these, shallow-rooted crops must include at least shallow-rooted fruits (such as bananas and pineapples) and other shallow-rooted crops (such as sisal and galangal). Medicinal herbs and shallow-rooted flowers grown in greenhouses cannot be effectively identified due to spectral obstruction and are therefore not included in the reference set.

[0040] Table 1. List of shallow-rooted crops that can be planted on arable land It should be noted that the data in Table 1 is based on the "List of Shallow-Root Crops that Can Be Planted on Cultivated Land in Guangdong Province (Trial Implementation in 2025)". In subsequent implementation, the crop list can be dynamically adjusted in combination with the actual development of agricultural production in the region.

[0041] For crops identified by remote sensing, a "field sampling + remote sensing synchronous acquisition" method was used to collect vegetation image features (leaf, flower, fruit shape, etc.), multi-band spectral reflectance features (mean, variance, peak, etc. of reflectance in blue, green, red, and near-infrared bands), vegetation index parameters (normalized vegetation index NDVI, enhanced vegetation index EVI, ratio vegetation index RVI, etc., see Table 2 for details) and phenological time-series features (time-series change curves within the growth cycle) during key phenological periods. These features were then associated with crop type codes to form a structured sample reference database.

[0042] Table 2 Vegetation Index Parameters Subsequently, a collection of intact topsoil patches was obtained. Multi-temporal, multispectral remote sensing images of the corresponding region (such as Sentinel-2 images) are required. These images must cover the key growth stages of crops, have a spatial resolution of at least 10 meters, and include blue, green, red, and near-infrared bands. Preprocessing is performed on the images, including radiometric calibration, atmospheric correction, geometric fine correction, and image fusion, to eliminate sensor and environmental errors and ensure data quality. Then, based on the preprocessed images, a set of intact topsoil patches is extracted. The vegetation feature parameters of each patch include vegetation image features, spectral reflectance features, vegetation index parameters, and phenological features.

[0043] Finally, the extracted feature parameters of the image patches are compared with the remote sensing identification sample reference set for similarity. A machine learning method with proven accuracy is used to calculate the feature vector matching degree, and an 85% confidence threshold is set. If the matching degree is higher than the confidence threshold If the value is below the confidence threshold, it is directly identified as the corresponding crop type; if the value is below the confidence threshold, it is directly identified as the corresponding crop type. If the image patch is located inside a greenhouse, it is marked as awaiting manual review. This review is conducted by professionals who combine high-resolution imagery, on-site survey photos, and other multi-source data for comprehensive confirmation. Finally, the automatic determination results and the manual review conclusions are combined to form a crop type identification result set containing complete crop type attributes. .

[0044] In step S105 of some embodiments, the contiguousness analysis refers to the analysis process of determining whether the area of ​​a patch itself or the combined area with surrounding related patches meets the requirements of large-scale agricultural production. The contiguousness optimization result set refers to the set of patches that meet the contiguousness conditions, adapting to the needs of large-scale farming.

[0045] For the patches in the crop type identification result set, combined with the current cultivated land and permanent basic farmland data, the area of ​​the patch itself and the combined area with the surrounding related patches are calculated and compared with the area threshold to conduct a concentrated contiguous analysis, screen out the patches suitable for large-scale utilization, and obtain the contiguous optimization result set.

[0046] In some embodiments, a contiguous analysis is performed on the patches in the crop type identification result set to obtain a contiguous optimization result set, including: For each patch in the crop type identification result set, the following judgments are performed sequentially: If the area of ​​the patch is greater than the first area threshold, it is determined to be a patch that meets the contiguous condition. Otherwise, if the combined area of ​​the patch and existing cultivated land and permanent basic farmland patches within its preset range is greater than the second area threshold, it is determined to be a patch that meets the contiguous condition. All map patches that are determined to meet the contiguous area conditions are compiled to form a contiguous area optimization result set.

[0047] Specifically, the first step is to prepare reference data. Based on the 2024 land change survey vector data, current cultivated land vector data for that year is obtained, along with permanent basic farmland data. These two types of data are used as references for determining the contiguousness of map patches. Subsequently, the crop type identification result set is analyzed. For each map patch, the contiguous area condition is determined sequentially: First, the area of ​​the map patch itself is calculated and compared with a first area threshold (e.g., 5 mu). If the area of ​​the map patch itself is greater than the threshold, it is directly determined to be a map patch that meets the contiguous area condition. Second, if the area of ​​the map patch itself does not reach the first area threshold, the combined area of ​​the map patch with existing cultivated land patches and permanent basic farmland patches within a preset range (e.g., 50 meters) is further calculated, and this combined area is compared with a second area threshold (e.g., 15 mu). If the combined area is greater than the threshold, it is determined to be a map patch that meets the contiguous area condition. Finally, all map patches that have been confirmed to meet the contiguous area condition through the above determination process are summarized to form a contiguous area optimization result set. This ensures that the selected image patches are suitable for the needs of large-scale agricultural production.

[0048] In step S106 of some embodiments, patch post-processing refers to processing the contiguous optimization result set. A series of optimization operations, including the removal of small land parcels, geometric correction, and confirmation of disposal information, are performed to improve the quality of the results. The cultivated land retroactive identification vector layer refers to a standardized vector layer that contains complete attributes of land parcels, including historical land type information, ownership information, cultivated layer status, crop type, disposal method, and disposal reason. It is the final result of cultivated land retroactive identification.

[0049] The post-processing of the contiguous optimization result set involves removing small patches, geometric correction, and confirming disposal information. It integrates the historical land type, ownership, topsoil status, crop type, disposal method, and other attributes of the patches to generate a cultivated land retroactive vector layer, forming a complete cultivated land retroactive result.

[0050] In some embodiments, post-processing of the contiguous optimization result set is performed to generate a cultivated land retrospective vector layer, including: Based on the contiguous optimization result set and the preset minimum area threshold, small patches with an area smaller than the preset minimum area threshold are removed to construct a clean patch set. Based on the cleaned patch set and preset geometric correction rules, the self-intersection topological error of the patch boundary is corrected, and the patch that meets the narrow and long judgment condition is fused or clipped to construct the geometrically corrected patch set. Based on the geometrically corrected map patch set, crop type identification results, topsoil integrity judgment conclusions, and multi-source auxiliary verification data, the disposal method and corresponding disposal reasons for each map patch are determined. Combining the historical land type information, ownership information, area data, topsoil status, crop type, disposal method, and disposal reasons for each map patch, a cultivated land retroactive vector layer is constructed and output.

[0051] Specifically, firstly, based on the contiguous optimization result set And a preset minimum area threshold (e.g., 1 acre), by calculating this result set The area of ​​all patches in the dataset is calculated, and small, scattered patches with areas smaller than a certain threshold are removed to complete the initial patch purification, thus constructing a purified patch set. .

[0052] Subsequently, based on the cleanup map set The system employs preset geometric correction rules to refine the data and construct a set of geometrically corrected patches. These rules include two types of requirements: self-intersection correction and narrow-length correction. Self-intersection correction addresses topological errors at patch boundaries by automatically deleting self-intersecting segments and reconstructing the patch boundaries using professional geographic information tools, ensuring topological integrity. Narrow-length correction uses a perimeter-to-area ratio > 0.5 as the criterion for narrow-length. For patches that meet this criterion, if the perimeter-to-area ratio can be reduced to 0.5 or below after merging with adjacent qualified patches (distance ≤ 5m), merging is performed. If merging is not possible, the narrow-length portion is trimmed (retaining a main body area ≥ 1 mu) to ensure that the patch shape adapts to agricultural production needs.

[0053] Finally, based on the geometrically corrected map patch set, combined with crop type identification results (including automatic and manual judgments), topsoil integrity assessment results, and multi-source auxiliary verification data (such as high-resolution imagery and field survey records), professionals determine the handling method and corresponding reasons for "recognizing historical outflow of cultivated land" or "not recognizing" each map patch. Then, by integrating the historical land use codes and names, ownership unit information, location information, map patch area, topsoil status, automatic judgment confidence level, and other attributes of the map patches, as well as the aforementioned handling methods and reasons, a cultivated land reclamation vector layer containing complete traceability information is constructed and output. Its attribute table must clearly cover key fields such as identification codes and element codes to ensure that the results are traceable and verifiable.

[0054] For example, the attribute table of the cultivated land retroactive vector layer includes at least the following fields: Identifier Code (BSM), Feature Code (YSDM), Land Category Code for 2019-2024 (DLBM_2019,...,DLBM_2024), Land Category Name for 2019-2024 (DLMC_2019,...,DLMC_2024), Ownership Unit Code (QSDWDM), Ownership Unit Name (QSDWMC), Location Unit Code (ZLDWDM), Location Unit Name The data includes: name (ZLDWMC), area of ​​the plot (TBMJ), state of the topsoil (GZCZT, assigned the values ​​"intact" and "damaged"), confidence level of automatic determination (ZDPDZXD), result of automatic determination (ZDPDJG, assigned the values ​​"specific crop type", "low confidence", and "located in greenhouse"), result of manual determination (RGCDJG, assigned the value "specific crop type"), crop type (ZWLX), disposal method (CZFS, assigned the values ​​"historical outflow of cultivated land" and "no re-recognition"), and reason for disposal (CZYY).

[0055] This method achieves precise tracing of historically outflowed arable land through multi-temporal land change survey vector data. It combines multiple spatial constraints with rigid integration of data assurance and planning control, relies on soil survey data to accurately determine the integrity of the topsoil, utilizes multispectral remote sensing data for efficient crop type identification, and selects suitable plots for large-scale utilization through concentrated contiguous analysis. Finally, it optimizes the quality of results through post-processing of map features, forming a complete technical process from historical tracing and multi-dimensional screening to result generation. This method not only achieves batch and precise identification of historically outflowed arable land, ensuring that the tracing results meet multiple requirements of spatial planning, cultivation conditions, crop types, and large-scale utilization, but also significantly reduces the subjectivity and workload of manual intervention, improving the efficiency, standardization, and reliability of arable land tracing work, providing solid technical support for the optimization and refined management of arable land resources.

[0056] Please see Figure 2 This application also provides a device for identifying and tracing historical outflow of cultivated land, which can implement the above-mentioned method for identifying and tracing historical outflow of cultivated land. The device includes: Module 201 is used to construct a candidate base map of historical outflow of cultivated land based on multi-time series land change survey vector data; The elimination module 202 is used to generate a constraint layer based on multiple spatial constraint data, and to perform spatial intersection elimination based on the historical outflow candidate base map and constraint layer to obtain the candidate set after avoidance. The judgment module 203 is used to judge the integrity of the topsoil based on the candidate set after avoidance and the soil survey data, and obtain the topsoil integrity patch set; The identification module 204 is used to identify crop types from the intact topsoil patches based on multispectral remote sensing data, and obtain a crop type identification result set. Analysis module 205 is used to perform concentrated contiguous analysis on the patches in the crop type identification result set to obtain a contiguous optimization result set; The generation module 206 is used to perform post-processing of the contiguous optimization result set to generate a cultivated land retrospective vector layer.

[0057] The specific implementation of this historical outflow farmland identification and retroactive recognition device is basically the same as the specific implementation of the above-mentioned historical outflow farmland identification and retroactive recognition method, and will not be repeated here.

[0058] Thirdly, embodiments of this application provide an electronic device, see [link to relevant documentation]. Figure 3 The diagram shown is a structural schematic of an electronic device provided in this application.

[0059] like Figure 3 As shown, the device includes: Memory 31 is used to store computer programs; Processor 32 is used to execute computer programs; When the processor 32 executes the computer program, it implements the historical outflow farmland identification and retroactive recognition method as described in any of the above embodiments.

[0060] For example, a computer program may be divided into one or more modules / units, one or more of which are stored in memory 31 and executed by processor 32 to complete this application. One or more modules / 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 an electronic device.

[0061] The processor 32 may 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 may be a microprocessor or any conventional processor.

[0062] The memory 31 can be used to store computer programs and / or modules. The processor 32 implements various functions of the electronic device by running or executing the computer programs and / or modules stored in the memory 31 and calling the data stored in the memory 31. The memory 31 may mainly include a program storage area and a data storage area. The program storage area may store the operating system, application programs required for at least one function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created according to the use of the mobile phone (such as audio data, phonebook, etc.). In addition, the memory 31 may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, RAM, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0063] It should be noted that the aforementioned electronic devices include, but are not limited to, processors and memory, as will be understood by those skilled in the art. Figure 3The structural diagram is merely an example of the electronic device described above and does not constitute a limitation on the electronic device. It may include more components than shown in the diagram, or combine certain components, or use different components.

[0064] Fourthly, embodiments of this application also provide a computer-readable storage medium storing a computer program that, when executed, implements the historical outflow farmland identification and retroactive recognition method of any of the above embodiments.

[0065] It should be understood that the implementation of all or part of the above-described method for identifying and retroactively recognizing historically outflowed arable land can also be accomplished 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 above-described method for identifying and retroactively recognizing historically outflowed arable land. 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 computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc. It should be noted that the content contained in the computer-readable medium can be appropriately added to or subtracted according to the requirements of legislation and patent practice in the relevant jurisdiction. For example, in some relevant jurisdictions, according to legislation and patent practice, the computer-readable medium does not include electrical carrier signals and telecommunication signals.

[0066] Fifthly, embodiments of this application also provide a computer program product, which is stored in a storage medium and executed by at least one processor to implement the historical outflow farmland identification and retroactive recognition method of any of the above embodiments.

[0067] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0068] The above description is the preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications are also considered to be within the scope of protection of this application.

Claims

1. A method for identifying and retroactively recognizing historically outflowed arable land, characterized in that, include: Based on multi-time series land change survey vector data, a candidate base map of historical outflow of cultivated land is constructed; A constraint layer is generated based on multiple spatial constraint data. Spatial intersection elimination is performed based on the historical outflow candidate base map and the constraint layer to obtain the candidate set after avoidance. Based on the candidate set after avoidance and the soil survey data, the integrity of the topsoil is judged to obtain a set of intact topsoil patches; Crop type identification is performed on the intact patch set of the topsoil layer based on multispectral remote sensing data to obtain a crop type identification result set; A concentrated contiguousness analysis was performed on the patches in the crop type identification result set to obtain a contiguous optimization result set; The contiguous optimization result set is subjected to patch post-processing to generate a cultivated land retrospective vector layer.

2. The method for identifying and retroactively recognizing historical outflow of arable land as described in claim 1, characterized in that, The construction of a candidate base map for historical outflow of cultivated land based on multi-time series land change survey vector data includes: Obtain vector data from land use change surveys for multiple consecutive years; Select map patches whose land type was cultivated land in the first base year; From the map patches, those whose land use type changed to orchard or woodland in any subsequent monitoring year are selected and output as candidate base maps of historical outflow of cultivated land.

3. The method for identifying and retroactively recognizing historical outflow of arable land as described in claim 1, characterized in that, The determination of the integrity of the topsoil layer requires that the following conditions be met simultaneously: The soil texture is within the preset suitable tillage range; The effective soil layer thickness shall not be less than the preset soil layer thickness threshold. The soil organic matter content shall not be lower than the lower limit of the background value of soil organic matter in the pre-defined area; The soil pH value is within the preset range suitable for crop growth.

4. The method for identifying and retroactively recognizing historical outflow of arable land as described in claim 1, characterized in that, The crop type identification is performed on the intact patch set of the topsoil based on multispectral remote sensing data to obtain a crop type identification result set, including: Construct a remote sensing identification sample reference set for shallow-rooted crops, wherein the shallow-rooted crops include at least shallow-rooted fruits and other shallow-rooted crop categories; Acquire multi-temporal multispectral remote sensing images of the region corresponding to the intact topsoil patch set and perform preprocessing; Extract vegetation feature parameters from the preprocessed remote sensing images; The extracted feature parameters are compared with the remote sensing identification sample reference set for similarity, and the crop type is determined according to the confidence threshold to obtain the crop type identification result set.

5. The method for identifying and retroactively recognizing historical outflow of arable land as described in claim 1, characterized in that, The analysis of the contiguous patches in the crop type identification result set yields a contiguous optimization result set, including: For each patch in the crop type identification result set, the following judgments are performed sequentially: If the area of ​​the patch is greater than the first area threshold, it is determined to be a patch that meets the contiguous condition. Otherwise, if the combined area of ​​the patch and existing cultivated land and permanent basic farmland patches within its preset range is greater than the second area threshold, it is determined to be a patch that meets the contiguous condition. All the image patches that are determined to meet the contiguous area conditions are summarized to form the contiguous area optimization result set.

6. The method for identifying and retroactively recognizing historical outflow of arable land as described in claim 1, characterized in that, The step of performing patch post-processing on the contiguous optimization result set to generate a cultivated land retrospective vector layer includes: Based on the contiguous optimization result set and the preset minimum area threshold, small patches with an area smaller than the preset minimum area threshold are removed to construct a clean patch set. Based on the cleaned patch set and preset geometric correction rules, the self-intersection topological error of the patch boundary is corrected, and the patch that meets the narrow and long judgment condition is fused or clipped to construct the geometrically corrected patch set. Based on the geometrically corrected map patch set, crop type identification results, topsoil integrity judgment conclusions, and multi-source auxiliary verification data, the disposal method and corresponding disposal reasons for each map patch are determined. Combining the historical land type information, ownership information, area data, topsoil status, crop type, disposal method, and disposal reasons for each map patch, a cultivated land retroactive vector layer is constructed and output.

7. A device for identifying and tracing historical outflow of arable land, characterized in that, include: The module is used to construct candidate base maps of historical outflow of cultivated land based on multi-time series land change survey vector data; The elimination module is used to generate a constraint layer based on multiple spatial constraint data, and to perform spatial intersection elimination based on the historical outflow candidate base map and the constraint layer to obtain a candidate set after avoidance. The judgment module is used to judge the integrity of the topsoil based on the candidate set after avoidance and the soil survey data, and obtain the topsoil integrity patch set; The identification module is used to identify crop types from the intact patch set of the topsoil layer based on multispectral remote sensing data, and obtain a crop type identification result set. The analysis module is used to perform a concentrated contiguous analysis on the patches in the crop type identification result set to obtain a contiguous optimization result set; The generation module is used to perform post-processing of the contiguous optimization result set to generate a cultivated land retrospective vector layer.

8. An electronic device, characterized in that, The system includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor, when executing the computer program, implements the historical outflow farmland identification and retroactive recognition method as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored computer program, wherein, when the computer program is executed, it controls the device on which the computer-readable storage medium is located to perform the historical outflow farmland identification and retroactive recognition method as described in any one of claims 1 to 6.

10. A computer program product, characterized in that, The computer program product includes a computer program or computer instructions, which, when executed by a processor, implement the historical outflow farmland identification and retroactive recognition method as described in any one of claims 1 to 6.

Images