GIS-based high-standard farmland visualization method and system
Through a high-standard farmland visualization method based on GIS, using the time series analysis of vegetation index and surface humidity, a plot adjacency diagram and a three-dimensional accessibility cost surface were constructed. Combined with evidence combination rules, this method solved the problem of ignoring the relationship between plots and the influence of terrain in traditional farmland evaluation, and achieved a more accurate assessment of farmland production capacity.
Patent Information
- Application Number
- CN202510945195.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-07-09
- Publication Date
- 2025-10-10
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
Traditional farmland evaluation methods are labor-intensive, costly, and highly subjective in the rapid and dynamic evaluation of large-scale areas. They are unable to reflect the stability of farmland production capacity and the interannual variation trend of soil fertility. They also ignore the spatial correlation between plots and the impact of complex terrain on mechanical access, leading to biased evaluation results.
By acquiring GIS data and using time series analysis of vegetation index and surface humidity, a plot adjacency diagram is constructed, the area agglomeration and three-dimensional accessibility cost surface are calculated, and multi-indicator fusion is performed using evidence combination rules to generate a high-standard farmland grade thematic map.
Accurately assessing the land's potential for large-scale operations and mechanized operating conditions solves the problem of neglecting the connections between plots and the impact of terrain in traditional methods, and improves the objectivity and reliability of the evaluation.
Smart Images
Figure CN120763269A_ABST
Abstract
Description
Technical Field
[0001] This application belongs to the field of GIS, and in particular relates to a high-standard farmland visualization method and system based on GIS. Background Art
[0002] In the process of promoting the construction of high-standard farmland, scientific, accurate, and efficient evaluation of the quality and potential of existing farmland is a prerequisite for rational planning, targeted investment, and effective supervision. Traditional farmland evaluation methods primarily rely on field surveys, sampling and testing, and statistical annual report data. While these methods can provide relatively detailed information within a small area, they are generally labor-intensive, costly, time-consuming, and subjective. They are particularly difficult to adapt to the needs of rapid and dynamic evaluation across large-scale areas. With the rapid development of remote sensing (RS) and geographic information system (GIS) technologies, it has become possible to utilize their macroscopic, rapid, and objective characteristics to evaluate farmland. Existing technologies typically use remote sensing imagery to extract surface parameters such as land use type and vegetation cover at a single point in time, or combine them with thematic data such as soil and hydrological data to establish an evaluation index system and use models such as weighted summation for comprehensive evaluation.
[0003] However, existing methods for constructing evaluation indicators often focus on static or single-temporal metrics, such as soil nutrient content or vegetation growth at a specific time. These methods fail to fully explore and utilize the dynamic information contained in multi-temporal remote sensing data, making it difficult to reflect the stability of farmland productivity and interannual trends in soil fertility. Furthermore, in spatial pattern analysis, most evaluation models treat each plot as an independent evaluation unit, ignoring the spatial connections between plots, such as their connectivity and agglomeration effects. These factors directly impact the efficiency of large-scale mechanized operations and the ease of organizing modern agricultural production. For the assessment of farming convenience, simple straight-line or network distances are often used to measure accessibility from plots to roads, with little consideration of the obstacles that complex terrain (such as slope) can pose to agricultural machinery. This can lead to discrepancies between accessibility evaluation results and actual conditions. For multi-indicator information fusion, the linear weighted model employed is somewhat subjective in weight determination and has limited ability to address potential conflicts and uncertainties between different indicators, compromising the objectivity and reliability of the evaluation results. Summary of the Invention
[0004] To address the problems existing in the prior art, this application proposes a high-standard farmland visualization method based on GIS, which includes the following steps: Obtaining GIS data for the area to be evaluated, the GIS data including at least two phases of remote sensing images, a digital elevation model (DEM), plot boundary vector data, road network vector data, and water system vector data; processing the at least two phases of remote sensing images to extract vegetation indices and surface humidity for each plot in each phase, and calculating the time series variation coefficient of the vegetation index and the time series stability of the surface humidity for each plot based on the extracted time series data; Based on the vector data of plot boundaries, road networks, and water systems, a plot adjacency graph was constructed. The graph uses plots as nodes and the spatial adjacency and infrastructure connectivity between plots as edges. The plot adjacency graph was used to calculate the regional agglomeration of each plot. A three-dimensional accessibility cost surface was generated using the road network vector data and DEM data, with terrain slope as the travel resistance. The average cumulative travel cost from each plot to the nearest road network was calculated. The temporal variation coefficient of the vegetation index, the temporal stability of surface humidity, the district agglomeration and the average cumulative travel cost are used as evaluation indicators, and the value of each indicator is mapped to the evidence credibility of the corresponding evaluation dimension. The evidence combination rules are used for fusion, and the comprehensive confidence and plausibility of each plot belonging to high-standard farmland are calculated; according to the comprehensive confidence and / or plausibility, each plot is graded and polygonal rendered to generate a high-standard farmland grade thematic map.
[0005] Preferably, the processing of the at least two periods of remote sensing images to extract the vegetation index and surface humidity of each plot in each period, and calculating the time series variation coefficient of the vegetation index and the time series stability of the surface humidity of each plot based on the extracted time series data, specifically includes: For each period of remote sensing images, the Normalized Difference Vegetation Index (NDVI) and the Normalized Difference Water Index (NDWI) are calculated respectively; For each plot, the mean NDVI and NDWI of all pixels within it are extracted to form the NDVI and NDWI time series of the plot; Based on the time series, the NDVI temporal variation coefficient and NDWI temporal stability of each plot were calculated.
[0006] Preferably, the method of calculating the area agglomeration degree of each plot using the plot adjacency relationship diagram specifically includes: Abstract each plot as a node in the graph; If two plots share a common boundary, a spatial adjacency edge is established between the corresponding nodes; If two plots are adjacent to the same road or water feature, an infrastructure connection edge is established between the corresponding nodes; The area agglomeration degree is the weighted sum of all spatial adjacent edges and infrastructure connected edges of a plot node according to preset weights.
[0007] Preferably, the method combines the road network vector data with the DEM data, uses the terrain slope as the travel resistance to generate a three-dimensional accessibility cost surface, and calculates the average cumulative travel cost from each block to the nearest road network, specifically including: Calculate the slope value of each grid cell based on the DEM data and set the passage cost according to the slope value, where the passage cost is set as a function that increases with the slope grade; Using the road network vector data as the starting source, the cost distance algorithm is used to generate the cost surface; For each plot, the cumulative travel cost values of all grids within it are extracted and the average value is calculated as the average cumulative travel cost of the plot.
[0008] Preferably, the fusion using evidence combination rules specifically includes: According to the positive or negative correlation between the value of each evaluation indicator and high-standard farmland, basic probabilities are assigned to the proposition "high-standard farmland" and global uncertainty in the identification framework {high-standard farmland, non-high-standard farmland}; Among them, for the surface moisture time series stability and area agglomeration, the larger the normalized value, the higher the basic probability of being assigned to "high-standard farmland"; for the vegetation index time series variation coefficient and average cumulative travel cost, the larger the normalized value, the lower the basic probability of being assigned to "high-standard farmland"; The Dempster combination rule is used to fuse the evidence generated by each indicator.
[0009] Preferably, the step of performing grade classification and polygon rendering on each plot based on the comprehensive trust and / or plausibility to generate a high-standard farmland grade thematic map specifically includes: Set up multi-level threshold rules based on comprehensive trust and comprehensive plausibility to classify land parcels into one of the following levels: "first-class", "second-class", "third-class" or "other-class"; Assign different colors or symbols to different levels of parcel polygons for visual rendering.
[0010] The present invention also proposes a high-standard farmland visualization system based on GIS, the system comprising: An extraction and calculation unit is used to obtain GIS data of the area to be evaluated, wherein the GIS data includes at least two periods of remote sensing images, a digital elevation model (DEM), plot boundary vector data, road network vector data, and water system vector data; process the at least two periods of remote sensing images, extract the vegetation index and surface humidity of each plot in each period, and calculate the time series variation coefficient of the vegetation index and the time series stability of the surface humidity of each plot based on the extracted time series data; The access calculation unit is used to construct a parcel adjacency diagram based on parcel boundaries, road network, and water system vector data. The diagram uses parcels as nodes and spatial adjacency and infrastructure connectivity between parcels as edges. The parcel adjacency diagram is used to calculate the area agglomeration of each parcel. The road network vector data and DEM data are combined to generate a three-dimensional accessibility cost surface with terrain slope as the travel resistance. The average cumulative travel cost from each parcel to the nearest road network is calculated. The grading and rendering unit is used to use the temporal variation coefficient of the vegetation index, the temporal stability of surface humidity, the district agglomeration and the average cumulative travel cost as evaluation indicators, and map each indicator value to the evidence credibility of the corresponding evaluation dimension, and use the evidence combination rule to fuse them, and calculate the comprehensive confidence and plausibility of each plot belonging to high-standard farmland; grade and polygon render each plot according to the comprehensive confidence and / or plausibility to generate a high-standard farmland grade thematic map.
[0011] Preferably, the processing of the at least two periods of remote sensing images to extract the vegetation index and surface humidity of each plot in each period, and calculating the time series variation coefficient of the vegetation index and the time series stability of the surface humidity of each plot based on the extracted time series data, specifically includes: For each period of remote sensing images, the Normalized Difference Vegetation Index (NDVI) and the Normalized Difference Water Index (NDWI) are calculated respectively; For each plot, the mean NDVI and NDWI of all pixels within it are extracted to form the NDVI and NDWI time series of the plot; Based on the time series, the NDVI temporal variation coefficient and NDWI temporal stability of each plot were calculated.
[0012] Preferably, the method of calculating the area agglomeration degree of each plot using the plot adjacency relationship diagram specifically includes: Abstract each plot as a node in the graph; If two plots share a common boundary, a spatial adjacency edge is established between the corresponding nodes; If two plots are adjacent to the same road or water feature, an infrastructure connection edge is established between the corresponding nodes; The area agglomeration degree is the weighted sum of all spatial adjacent edges and infrastructure connected edges of a plot node according to preset weights.
[0013] Preferably, the method combines the road network vector data with the DEM data, uses the terrain slope as the travel resistance to generate a three-dimensional accessibility cost surface, and calculates the average cumulative travel cost from each block to the nearest road network, specifically including: Calculate the slope value of each grid cell based on the DEM data and set the passage cost according to the slope value, where the passage cost is set as a function that increases with the slope grade; Using the road network vector data as the starting source, the cost distance algorithm is used to generate the cost surface; For each plot, the cumulative travel cost values of all grids within it are extracted and the average value is calculated as the average cumulative travel cost of the plot.
[0014] Preferably, the fusion using evidence combination rules specifically includes: According to the positive or negative correlation between the value of each evaluation indicator and high-standard farmland, basic probabilities are assigned to the proposition "high-standard farmland" and global uncertainty in the identification framework {high-standard farmland, non-high-standard farmland}; Among them, for the surface moisture time series stability and area agglomeration, the larger the normalized value, the higher the basic probability of being assigned to "high-standard farmland"; for the vegetation index time series variation coefficient and average cumulative travel cost, the larger the normalized value, the lower the basic probability of being assigned to "high-standard farmland"; The Dempster combination rule is used to fuse the evidence generated by each indicator.
[0015] Preferably, the step of performing grade classification and polygon rendering on each plot based on the comprehensive trust and / or plausibility to generate a high-standard farmland grade thematic map specifically includes: Set up multi-level threshold rules based on comprehensive trust and comprehensive plausibility to classify land parcels into one of the following levels: "first-class", "second-class", "third-class" or "other-class"; Assign different colors or symbols to different levels of parcel polygons for visual rendering.
[0016] By introducing the time series analysis of vegetation index and surface humidity, the present invention can reveal the stability and sustainability of farmland production capacity from a temporal dimension, overcoming the defect of traditional methods that rely only on static or single-phase indicators and cannot fully reflect the soil conditions. By constructing a plot adjacency relationship diagram to quantify the area agglomeration, and combining the terrain slope to calculate the actual travel cost, it can more accurately evaluate the scale operation potential and mechanized operation conditions of the plot from the two levels of spatial structure and actual farming convenience, making up for the shortcomings of existing technologies that ignore the correlation between plots and the influence of terrain. In addition, the use of evidence combination rules for multi-indicator fusion solves the uncertainty and conflict problems between information from different sources and avoids the subjective drawbacks of weight setting in linear weighted models. BRIEF DESCRIPTION OF THE DRAWINGS
[0017] Figure 1 This is a flow chart of Example 1; Figure 2 This is a schematic diagram of the changes in the time series indicators of the plot; Figure 3 Schematic diagram of plot adjacency and agglomeration; Figure 4 Schematic diagram of fusion using evidence combination rules. DETAILED DESCRIPTION
[0018] 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 embodiments described 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 those skilled in the art without making creative efforts are within the scope of protection of this application.
[0019] In a specific embodiment, the present application proposes a high-standard farmland visualization method based on GIS, comprising the following steps: S1, obtaining GIS data of the area to be evaluated, the GIS data including at least two periods of remote sensing images, a digital elevation model (DEM), plot boundary vector data, road network vector data, and water system vector data; processing the at least two periods of remote sensing images, extracting vegetation index and surface humidity for each plot in each period, and calculating the time series variation coefficient of vegetation index and time series stability of surface humidity for each plot based on the extracted time series data; Remote sensing images were collected from Landsat satellite images covering the area under evaluation. Cloudless or lightly cloudy images were selected during the main crop growing season of each year over the past five years, such as from May to October. The Digital Elevation Model (DEM) was constructed using the Shuttle Radar Topography Mission (SRTM) data. Vector data for plot boundaries, road networks, and water systems were obtained from the latest land survey databases or urban and rural planning authorities. Each period of remote sensing imagery was preprocessed with radiometric calibration and atmospheric correction. The Normalized Difference Vegetation Index (NDVI) for each period was calculated using the formula NIR-Red / NIR+Red using the near-infrared and red bands. The surface moisture component was extracted using the tasseled cap transform. Within each plot, the mean NDVI and surface moisture mean series for each corresponding period were extracted over multiple years. The coefficient of variation (CV) for the NDVI series for each plot was calculated using the standard deviation divided by the mean. Similarly, the CV of the surface moisture series was calculated. A smaller CV value indicates greater temporal stability.
[0020] In an optional embodiment, the processing of the at least two periods of remote sensing images to extract the vegetation index and surface humidity of each plot in each period, and calculating the time series variation coefficient of the vegetation index and the time series stability of the surface humidity of each plot based on the extracted time series data, specifically includes: For each period of remote sensing images, the Normalized Difference Vegetation Index (NDVI) and the Normalized Difference Water Index (NDWI) are calculated respectively; For each plot, the mean NDVI and NDWI of all pixels within it are extracted to form the NDVI and NDWI time series of the plot; Based on the time series, the NDVI temporal variation coefficient and NDWI temporal stability of each plot were calculated.
[0021] Assume that three Sentinel-2 remote sensing images are acquired for the study area during key crop growing seasons of the year, such as May, July, and September. For each image, NDVI is calculated using the near-infrared and red bands, and NDWI is calculated using the green and near-infrared bands.
[0022] For a specific farmland plot A, extract the NDVI values of all pixels within its boundary from these three image periods and calculate the average, which may result in a time series such as 0.3, 0.8, and 0.5. Similarly, calculate the NDWI time series for plot A, which may be 0.2, 0.1, and 0.15.
[0023] Based on these time series, the characteristic indicators of the plots are further calculated. For the NDVI time series 0.3, 0.8, and 0.5, the mean is 0.53, the standard deviation is about 0.25, and the calculated time series coefficient of variation is 0.47. A lower coefficient of variation, such as less than 0.2, usually indicates that the vegetation cover changes smoothly and continuously at high values during the crop growth cycle, which is a characteristic of well-managed high-standard farmland. For the NDWI time series, its time series stability is calculated. The stability can be measured in a way opposite to the coefficient of variation, such as by subtracting the coefficient of variation from 1. An NDWI time series stability value close to 1 indicates that the moisture conditions of the plot are very stable throughout the growing period, which means that it may have a reliable irrigation system. Figure 2 A schematic diagram showing the changes in the time series values of the indicator index NDVI and the time series values of surface humidity from May to September is shown.
[0024] S2: Based on the vector data of plot boundaries, road networks, and water systems, a plot adjacency graph is constructed. The graph uses plots as nodes and the spatial adjacency and infrastructure connectivity between plots as edges. The plot adjacency graph is used to calculate the regional agglomeration of each plot. A three-dimensional accessibility cost surface is generated using the road network vector data and DEM data, with terrain slope as the travel resistance. The average cumulative travel cost from each plot to the nearest road network is calculated. The geometric center of each plot is abstracted as a node in the graph; if two plot polygons share a boundary, an edge is established between the corresponding nodes; if two plots are completely separated by a road or a canal but do not touch each other, they are judged to be connected by infrastructure, and an edge is also established between the corresponding nodes. Figure 3 A schematic diagram shows the spatial adjacency and facility connectivity between different plots. Once the graph is constructed, a degree centrality algorithm is used to calculate the degree of each node, which is the number of edges directly connected to the node. This degree value is the district agglomeration of the plot.
[0025] A slope raster map was generated using DEM data. The slope values were then reclassified into travel cost values based on the agricultural machinery's traffic capacity, such as a slope less than 5 degrees was set to a low cost of 1, a slope between 5 and 15 degrees was set to a medium cost of 5, and a slope greater than 15 degrees was set to a high cost of 50. The road network vector data was used as the source and the slope cost raster was used as the resistance surface. The Dijkstra algorithm was used to calculate the minimum cumulative travel cost from each pixel in the area to the nearest road to generate a cost surface raster. The zonal statistics tool was used to calculate the average cost value of all pixels within the vector polygon of each plot, which was used as the average cumulative travel cost of the plot.
[0026] In an optional embodiment, the method of calculating the area agglomeration degree of each plot using the plot adjacency graph specifically includes: Abstract each plot as a node in the graph; If two plots share a common boundary, a spatial adjacency edge is established between the corresponding nodes; If two plots are adjacent to the same road or water feature, an infrastructure connection edge is established between the corresponding nodes; The area agglomeration degree is the weighted sum of all spatial adjacent edges and infrastructure connected edges of a plot node according to preset weights.
[0027] In a geographic information system, each farmland parcel polygon is treated as a node in a graph. For example, if Plot A and Plot B share a 200-meter-long common boundary, a spatial adjacency edge is connected between the nodes representing them. Similarly, if Plot A and Plot C, while not directly connected, are adjacent to the same tractor road, an infrastructure connectivity edge is connected between their corresponding nodes. This graph structure comprehensively describes the spatial layout and functional connections between the parcels.
[0028] To quantify the closeness of the connection, weights are assigned to different types of edges. For example, the weight of a spatially adjacent edge can be set to 1.0 because it represents the most direct physical connection, while the weight of an infrastructure-connected edge connected by a road or canal can be set to 0.7 because it represents functional connectivity. For plot A, if it is spatially adjacent to three plots and connected to two other plots by road, then its area agglomeration is 4.4. A high agglomeration value, such as greater than 5, means that the plot is closely surrounded by other farmland and infrastructure, forming a pattern of contiguous development, which is exactly the goal pursued by the planning and construction of high-standard farmland.
[0029] In an optional embodiment, the method combines the road network vector data with the DEM data, uses the terrain slope as the travel resistance to generate a three-dimensional accessibility cost surface, and calculates the average cumulative travel cost from each block to the nearest road network, specifically including: Calculate the slope value of each grid cell based on the DEM data and set the passage cost according to the slope value, where the passage cost is set as a function that increases with the slope grade; Using the road network vector data as the starting source, the cost distance algorithm is used to generate the cost surface; For each plot, the cumulative travel cost values of all grids within it are extracted and the average value is calculated as the average cumulative travel cost of the plot.
[0030] Using 10-meter resolution digital elevation model (DEM) data, the slope of each 10-meter by 10-meter grid in the study area was calculated. Next, a cost function was defined. For example, a flat area with a slope less than 2 degrees has a travel cost of 1; a gentle slope between 2 and 6 degrees has a cost of 5; and an area with a slope greater than 6 degrees has a cost of 50. The cost value represents the relative difficulty of agricultural machinery or personnel crossing the grid. The cost distance algorithm is run using the road network of all levels as the starting point with a cost of zero. The algorithm starts from the road and calculates the minimum cumulative travel cost required to reach each grid outward. The cumulative cost of a grid close to the road and with flat terrain may be in the single digits, such as 8. On the other hand, the cumulative cost of a grid far from the road and located on a steep hillside may be as high as thousands, such as 3,500. For a specific plot, the cumulative travel cost values of all the grids within it are added together and averaged. Plots with an average cumulative travel cost of less than 100 mean that their internal and surrounding transportation is extremely convenient, mechanized operations are highly efficient, and they are more in line with the requirements of high-standard farmland.
[0031] S3, taking the temporal variation coefficient of the vegetation index, the temporal stability of surface humidity, the district agglomeration and the average cumulative travel cost as evaluation indicators, and mapping each indicator value to the evidence credibility of the corresponding evaluation dimension, adopting the evidence combination rule to fuse them, and calculating the comprehensive confidence and plausibility of each plot belonging to high-standard farmland; according to the comprehensive confidence and / or plausibility, classify and polygon render each plot, and generate a high-standard farmland grade thematic map.
[0032] The evaluation identification framework is set as high-standard farmland and non-high-standard farmland; then, S-type or Z-type membership functions are established for the four evaluation indicators respectively, and the normalized values of each indicator are converted into basic probability distribution values for high-standard farmland and non-high-standard farmland, namely, evidence credibility; subsequently, the Dempster evidence combination rule is used to iteratively fuse the evidence credibility from the four indicators in pairs until a comprehensive basic probability distribution function is obtained; based on the comprehensive results, the trust function Bel value and likelihood function Pl value of each plot belonging to high-standard farmland are calculated.
[0033] Exemplarily, a grade classification rule is set. For example, when the trust degree Bel is greater than 0.7, it is classified as excellent; when the trust degree Bel is between 0.4 and 0.7 and the plausibility Pl is greater than 0.6, it is classified as good; when the trust degree Bel is less than 0.4, it is classified as medium; the rest are classified as poor; in GIS software, the four grades of excellent, good, medium and poor are assigned dark green, light green, yellow and red respectively, and the polygons of each block are filled with colors according to their evaluation levels, and a visual thematic map is output.
[0034] In an optional embodiment, the fusion using evidence combination rules specifically includes: According to the positive or negative correlation between the value of each evaluation indicator and high-standard farmland, basic probabilities are assigned to the proposition "high-standard farmland" and global uncertainty in the identification framework {high-standard farmland, non-high-standard farmland}; Among them, for the surface moisture time series stability and area agglomeration, the larger the normalized value, the higher the basic probability of being assigned to "high-standard farmland"; for the vegetation index time series variation coefficient and average cumulative travel cost, the larger the normalized value, the lower the basic probability of being assigned to "high-standard farmland"; The Dempster combination rule is used to fuse the evidence generated by each indicator.
[0035] All indicator values are normalized to the range of 0 to 1. For a plot, if its normalized area agglomeration value is 0.9, based on its positive correlation, a basic probability of 0.85 can be assigned to the proposition that it is high-standard farmland, and no probability can be assigned to the proposition that it is not high-standard farmland. The remaining 0.15 is allocated to global uncertainty, indicating that we are not completely confident based on this single piece of evidence.
[0036] On the contrary, if the normalized value of the average cumulative travel cost of the plot is 0.7, based on its negative correlation, only a basic probability of 0.1 may be assigned to the proposition high-standard farmland, while 0.6 is assigned to the proposition non-high-standard farmland, and the remaining 0.3 is assigned to global uncertainty. The Dempster combination rule provides a mathematical formula for combining the basic probabilities from four independent sources of evidence: area agglomeration, travel cost, vegetation index variability, and humidity stability. By combining them one by one, a comprehensive basic probability distribution is finally obtained, which takes into account all positive and negative information at the same time, thereby giving a more reliable comprehensive judgment on whether the plot is high-standard farmland than any single indicator. Figure 4 Schematic diagram of fusion using evidence combination rules.
[0037] In an optional embodiment, the step of performing grade classification and polygon rendering on each plot based on the comprehensive trust and / or plausibility to generate a high-standard farmland grade thematic map specifically includes: Set up multi-level threshold rules based on comprehensive trust and comprehensive plausibility to classify land parcels into one of the following levels: "first-class", "second-class", "third-class" or "other-class"; Assign different colors or symbols to different levels of parcel polygons for visual rendering.
[0038] After integrating all indicators through evidence theory, a comprehensive confidence and comprehensive plausibility score for the proposition "high-standard farmland" were obtained for each plot. For example, plot A has a comprehensive confidence score of 0.88 and a plausibility score of 0.95. For example, the following grading rules can be set: confidence scores greater than 0.8 are classified as first-class; confidence scores between 0.6 and 0.8 are classified as second-class; confidence scores between 0.4 and 0.6 are classified as third-class; and confidence scores less than 0.4 are classified as inferior. Based on these rules, plot A would be rated first-class.
[0039] Another plot, B, has a final confidence score of 0.52 and a plausibility score of 0.7, placing it at third grade. During the final map visualization phase, these grades are given distinct visual symbols. For example, all first-grade plots are filled in dark green, second-grade plots in light green, third-grade plots in yellow, and substandard plots in red. This allows for rapid identification of core areas of high-quality, high-standard farmland, areas requiring improvement, and substandard plots requiring focused remediation.
[0040] In an optional embodiment, before calculating metrics for each plot, the Digital Elevation Model (DEM) and remote sensing imagery from each phase are masked and partitioned based on the plot boundary vector data to ensure that all calculations are performed within the plot unit. The underlying data includes remote sensing imagery and Digital Elevation Model (DEM) data covering the entire project area, as well as a precise file of plot boundary vectors. Masking is performed before calculating any plot-level metrics, such as average NDVI or average slope. For example, for plot T001, its polygonal boundary is used to precisely crop the pixel region belonging to that plot from the entire remote sensing image, forming an irregular image subset. All pixel information outside the plot boundary is ignored. Suppose that plot T001 is inherently flat but is adjacent to the foot of a steep mountain. Without masking, the pixels at the foot of the mountain would be incorrectly included when calculating the average slope of this area, resulting in an inflated slope value that could potentially misclassify the plot as unsuitable for mechanized cultivation. By performing this masking and zoning statistics based on the plot boundaries for each period of remote sensing images and DEM data, it is ensured that every indicator value calculated subsequently, whether it is vegetation index, moisture index or terrain parameter, strictly and uniquely reflects the actual situation inside the plot.
[0041] Specific embodiment 2 proposes a high-standard farmland visualization system based on GIS, the system comprising: An extraction and calculation unit is used to obtain GIS data of the area to be evaluated, wherein the GIS data includes at least two periods of remote sensing images, a digital elevation model (DEM), plot boundary vector data, road network vector data, and water system vector data; process the at least two periods of remote sensing images, extract the vegetation index and surface humidity of each plot in each period, and calculate the time series variation coefficient of the vegetation index and the time series stability of the surface humidity of each plot based on the extracted time series data; The access calculation unit is used to construct a parcel adjacency diagram based on parcel boundaries, road network, and water system vector data. The diagram uses parcels as nodes and spatial adjacency and infrastructure connectivity between parcels as edges. The parcel adjacency diagram is used to calculate the area agglomeration of each parcel. The road network vector data and DEM data are combined to generate a three-dimensional accessibility cost surface with terrain slope as the travel resistance. The average cumulative travel cost from each parcel to the nearest road network is calculated. The grading and rendering unit is used to use the temporal variation coefficient of the vegetation index, the temporal stability of surface humidity, the district agglomeration and the average cumulative travel cost as evaluation indicators, and map each indicator value to the evidence credibility of the corresponding evaluation dimension, and use the evidence combination rule to fuse them, and calculate the comprehensive confidence and plausibility of each plot belonging to high-standard farmland; grade and polygon render each plot according to the comprehensive confidence and / or plausibility to generate a high-standard farmland grade thematic map.
[0042] In an optional embodiment, the processing of the at least two periods of remote sensing images to extract the vegetation index and surface humidity of each plot in each period, and calculating the time series variation coefficient of the vegetation index and the time series stability of the surface humidity of each plot based on the extracted time series data, specifically includes: For each period of remote sensing images, the Normalized Difference Vegetation Index (NDVI) and the Normalized Difference Water Index (NDWI) are calculated respectively; For each plot, the mean NDVI and NDWI of all pixels within it are extracted to form the NDVI and NDWI time series of the plot; Based on the time series, the NDVI temporal variation coefficient and NDWI temporal stability of each plot were calculated.
[0043] In an optional embodiment, the method of calculating the area agglomeration degree of each plot using the plot adjacency graph specifically includes: Abstract each plot as a node in the graph; If two plots share a common boundary, a spatial adjacency edge is established between the corresponding nodes; If two plots are adjacent to the same road or water feature, an infrastructure connection edge is established between the corresponding nodes; The area agglomeration degree is the weighted sum of all spatial adjacent edges and infrastructure connected edges of a plot node according to preset weights.
[0044] In an optional embodiment, the method combines the road network vector data with the DEM data, uses the terrain slope as the travel resistance to generate a three-dimensional accessibility cost surface, and calculates the average cumulative travel cost from each block to the nearest road network, specifically including: Calculate the slope value of each grid cell based on the DEM data and set the passage cost according to the slope value, where the passage cost is set as a function that increases with the slope grade; Using the road network vector data as the starting source, the cost distance algorithm is used to generate the cost surface; For each plot, the cumulative travel cost values of all grids within it are extracted and the average value is calculated as the average cumulative travel cost of the plot.
[0045] In an optional embodiment, the fusion using evidence combination rules specifically includes: According to the positive or negative correlation between the value of each evaluation indicator and high-standard farmland, basic probabilities are assigned to the proposition "high-standard farmland" and global uncertainty in the identification framework {high-standard farmland, non-high-standard farmland}; Among them, for the surface moisture time series stability and area agglomeration, the larger the normalized value, the higher the basic probability of being assigned to "high-standard farmland"; for the vegetation index time series variation coefficient and average cumulative travel cost, the larger the normalized value, the lower the basic probability of being assigned to "high-standard farmland"; The Dempster combination rule is used to fuse the evidence generated by each indicator.
[0046] In an optional embodiment, the step of performing grade classification and polygon rendering on each plot based on the comprehensive trust and / or plausibility to generate a high-standard farmland grade thematic map specifically includes: Set up multi-level threshold rules based on comprehensive trust and comprehensive plausibility to classify land parcels into one of the following levels: "first-class", "second-class", "third-class" or "other-class"; Assign different colors or symbols to different levels of parcel polygons for visual rendering.
[0047] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them. Although the present invention has been described in detail with reference to the above embodiments, it should be understood by those skilled in the art that the technical solutions described in the above embodiments can still be modified, or some of the technical features thereof can be replaced by equivalents. However, these modifications or replacements do not deviate from the spirit and scope of the technical solutions of the embodiments of the present invention. In addition, the various different implementations of the embodiments of the present invention can also be arbitrarily combined, as long as they do not violate the ideas of the embodiments of the present invention, and they should also be regarded as the contents disclosed in the embodiments of the present invention.
Claims
1. A high-standard farmland visualization method based on GIS, characterized in that: The following steps are involved: Obtaining GIS data for the area to be evaluated, the GIS data including at least two phases of remote sensing images, a digital elevation model (DEM), plot boundary vector data, road network vector data, and water system vector data; processing the at least two phases of remote sensing images to extract vegetation indices and surface humidity for each plot in each phase, and calculating the time series variation coefficient of the vegetation index and the time series stability of the surface humidity for each plot based on the extracted time series data; Based on the vector data of plot boundaries, road networks, and water systems, a plot adjacency graph was constructed. The graph uses plots as nodes and the spatial adjacency and infrastructure connectivity between plots as edges. The plot adjacency graph was used to calculate the regional agglomeration of each plot. A three-dimensional accessibility cost surface was generated using the road network vector data and DEM data, with terrain slope as the travel resistance. The average cumulative travel cost from each plot to the nearest road network was calculated. The vegetation index time series variation coefficient, surface humidity time series stability, area agglomeration, and average cumulative travel cost are used as evaluation indicators, and each indicator value is mapped to the evidence credibility of the corresponding evaluation dimension. The evidence combination rule is used for fusion to calculate the comprehensive confidence and plausibility of each plot belonging to high-standard farmland; Each block is graded and polygon rendered according to the comprehensive confidence and / or plausibility to generate a high-standard farmland grade thematic map.
2. The method according to claim 1, characterized in that The processing of the at least two periods of remote sensing images to extract the vegetation index and surface humidity of each plot in each period, and calculating the time series variation coefficient of the vegetation index and the time series stability of the surface humidity of each plot based on the extracted time series data, specifically includes: For each period of remote sensing images, the Normalized Difference Vegetation Index (NDVI) and the Normalized Difference Water Index (NDWI) are calculated respectively; For each plot, the mean NDVI and NDWI of all pixels within it are extracted to form the NDVI and NDWI time series of the plot; Based on the time series, the NDVI temporal variation coefficient and NDWI temporal stability of each plot were calculated.
3. The method according to claim 1, characterized in that The calculation of the area agglomeration of each block using the block adjacency relationship diagram specifically includes: Abstract each plot as a node in the graph; If two plots share a common boundary, a spatial adjacency edge is established between the corresponding nodes; If two plots are adjacent to the same road or water feature, an infrastructure connection edge is established between the corresponding nodes; The area agglomeration degree is the weighted sum of all spatial adjacent edges and infrastructure connected edges of a plot node according to preset weights.
4. The method according to claim 1, wherein The above method combines the road network vector data with the DEM data, uses the terrain slope as the travel resistance, and generates a three-dimensional accessibility cost surface. The average cumulative travel cost from each block to the nearest road network is calculated, specifically including: Calculate the slope value of each grid cell based on the DEM data, and set the passage cost according to the slope value, where the passage cost is set as a function that increases with the slope grade; Using the road network vector data as the starting source, the cost distance algorithm is used to generate the cost surface; For each plot, the cumulative travel cost values of all grids within it are extracted and the average value is calculated as the average cumulative travel cost of the plot.
5. The method according to claim 1, wherein The fusion using evidence combination rules specifically includes: According to the positive or negative correlation between the value of each evaluation indicator and high-standard farmland, basic probabilities are assigned to the proposition "high-standard farmland" and global uncertainty in the identification framework {high-standard farmland, non-high-standard farmland}; Among them, for the surface moisture time series stability and area agglomeration, the larger the normalized value, the higher the basic probability of being assigned to "high-standard farmland"; for the vegetation index time series variation coefficient and average cumulative travel cost, the larger the normalized value, the lower the basic probability of being assigned to "high-standard farmland"; The Dempster combination rule is used to fuse the evidence generated by each indicator.
6. The method according to claim 1, characterized in that The step of classifying and polygon rendering each block according to the comprehensive trust and / or plausibility to generate a high-standard farmland grade thematic map specifically includes: Set up multi-level threshold rules based on comprehensive trust and comprehensive plausibility to classify land parcels into one of the following levels: "first-class", "second-class", "third-class" or "other-class"; Assign different colors or symbols to different levels of parcel polygons for visual rendering.
7. A high-standard farmland visualization system based on GIS, characterized by: The system comprises: An extraction and calculation unit is used to obtain GIS data of the area to be evaluated, wherein the GIS data includes at least two periods of remote sensing images, a digital elevation model (DEM), plot boundary vector data, road network vector data, and water system vector data; process the at least two periods of remote sensing images, extract the vegetation index and surface humidity of each plot in each period, and calculate the time series variation coefficient of the vegetation index and the time series stability of the surface humidity of each plot based on the extracted time series data; The access calculation unit is used to construct a parcel adjacency diagram based on parcel boundaries, road network, and water system vector data. The diagram uses parcels as nodes and spatial adjacency and infrastructure connectivity between parcels as edges. The parcel adjacency diagram is used to calculate the area agglomeration of each parcel. The road network vector data and DEM data are combined to generate a three-dimensional accessibility cost surface with terrain slope as the travel resistance. The average cumulative travel cost from each parcel to the nearest road network is calculated. The grading and rendering unit is used to use the temporal variation coefficient of the vegetation index, the temporal stability of surface humidity, the district agglomeration and the average cumulative travel cost as evaluation indicators, and map each indicator value to the evidence credibility of the corresponding evaluation dimension, and use the evidence combination rule to fuse them, and calculate the comprehensive confidence and plausibility of each plot belonging to high-standard farmland; grade and polygon render each plot according to the comprehensive confidence and / or plausibility to generate a high-standard farmland grade thematic map.
8. The system according to claim 7, characterized in that The processing of the at least two periods of remote sensing images to extract the vegetation index and surface humidity of each plot in each period, and calculating the time series variation coefficient of the vegetation index and the time series stability of the surface humidity of each plot based on the extracted time series data, specifically includes: For each period of remote sensing images, the Normalized Difference Vegetation Index (NDVI) and the Normalized Difference Water Index (NDWI) are calculated respectively; For each plot, the mean NDVI and NDWI of all pixels within it are extracted to form the NDVI and NDWI time series of the plot; Based on the time series, the NDVI temporal variation coefficient and NDWI temporal stability of each plot were calculated.
9. The system according to claim 7, wherein: The calculation of the area agglomeration of each block using the block adjacency relationship diagram specifically includes: Abstract each plot as a node in the graph; If two plots share a common boundary, a spatial adjacency edge is established between the corresponding nodes; If two plots are adjacent to the same road or water feature, an infrastructure connection edge is established between the corresponding nodes; The area agglomeration degree is the weighted sum of all spatial adjacent edges and infrastructure connected edges of a plot node according to preset weights.
10. The system according to claim 7, wherein: The above method combines the road network vector data with the DEM data, uses the terrain slope as the travel resistance, and generates a three-dimensional accessibility cost surface. The average cumulative travel cost from each block to the nearest road network is calculated, specifically including: Calculate the slope value of each grid cell based on the DEM data, and set the passage cost according to the slope value, where the passage cost is set as a function that increases with the slope grade; Using the road network vector data as the starting source, the cost distance algorithm is used to generate the cost surface; For each plot, the cumulative travel cost values of all grids within it are extracted and the average value is calculated as the average cumulative travel cost of the plot.