CN-122022078-A - Permanent basic farmland intelligent repair and scratch management method and system based on data processing

CN122022078ACN 122022078 ACN122022078 ACN 122022078ACN-122022078-A

Abstract

The invention discloses a permanent basic farmland intelligent repair and scratch management method and system based on data processing, which belong to the field of geographic information systems, wherein the method comprises the steps of acquiring a multisource geographic space data set of a target area, mapping the multisource geographic space data set into a unified grid system, and constructing a multidimensional feature vector of each grid unit; based on the multidimensional feature vector, calculating the cultivated land suitability score of each grid unit by using a cultivated land suitability evaluation model comprising a space smoothness constraint mechanism to generate a cultivated land suitability distribution map, polymerizing the discrete grid units into a plurality of preliminary continuous plots by using an anisotropic region growth algorithm, and screening an optimal complementary scheme from the continuous plots by taking the complementary area as a rigid constraint. The invention is beneficial to forming a high-standard permanent basic farmland with concentrated continuous sheets.

Inventors

Deng Yunsi
WU LIJUN
LIU FAN
YANG BO
LI JUN
MU YU
WANG WEIWEI
LIU CONGXIN

Assignees

四川省核地质调查研究所
四川省国土整治中心

Dates

Publication Date: 20260512
Application Date: 20260413

Claims (10)

1. The permanent basic farmland intelligent repair and scratch management method based on data processing is characterized by comprising the following steps of: s100, acquiring a multi-source geospatial data set of a target area, mapping the multi-source geospatial data set into a unified grid system, and constructing a multi-dimensional feature vector of each grid unit; S200, calculating the cultivated land suitability score of each grid unit by using a cultivated land suitability evaluation model comprising a space smoothness constraint mechanism based on the multidimensional feature vector, and generating a cultivated land suitability distribution map; s300, based on the cultivated land suitability distribution map, utilizing an anisotropic region growing algorithm to aggregate the discrete grid units into a plurality of preliminary continuous plots, wherein, The anisotropic region growing algorithm fuses the terrain gradient direction constraint in the growing process; S400, constructing a multi-objective optimization function comprising a cultivated land quality maximization target and a space crushing degree minimization target by taking the complement area as a rigid constraint, and screening an optimal complement scheme from the preliminary continuous land parcels.
2. The permanent basic farmland intelligent repair and scratch management method based on data processing according to claim 1, wherein in S100, the construction process of the multidimensional feature vector includes: Dividing the target area into grid cells which are arranged regularly; Extracting multisource geospatial data corresponding to each grid unit, wherein the data at least comprise high-resolution remote sensing image data reflecting vegetation coverage characteristics, digital elevation model data reflecting three-dimensional terrain characteristics, soil physicochemical property data reflecting soil indicators and land utilization status and planning vector data reflecting space management and control states; and arranging the data according to a preset sequence to form the multidimensional feature vector, and carrying out dimensionless standardization processing on continuous variables in the multidimensional feature vector.
3. The permanent basic farmland intelligent repair and planning management method based on data processing according to claim 2, characterized in that the standardized processing specifically adopts extremely bad mapping logic: and calculating the relative position proportion of the characteristic value of the current grid unit in the value interval by counting the maximum value and the minimum value of the variable in the global range, and linearly mapping the physical characteristic to the [0,1] closed interval.
4. The permanent basic farmland intelligent repair and planning management method based on data processing according to claim 1, wherein the arable land suitability evaluation model in S200 includes: calculating basic scores, namely carrying out weighted accumulation on the standardized characteristic values of the grid cells according to the weight coefficients set by the expert knowledge base to obtain initial scores; rigid constraint filtering, namely judging whether the grid cells fall into an ecological protection red line or a forbidden construction area of town development boundaries, If yes, the initial score is forced to be zero by utilizing a binary constraint function, If not, the initial score is reserved; Space neighborhood correction, namely constructing a space neighborhood window centering on the current grid cell, calculating the average value of the basic suitability scores of all neighbor grid cells in the window, And weighting and summing the self basic score and the average value by using a preset smoothing coefficient to obtain a final cultivated land suitability score.
5. The permanent basic farmland intelligent repair and scratch management method based on data processing according to claim 1, wherein in S300, the anisotropic region growing algorithm comprises: Setting a suitability scoring threshold value, screening grid cells with scores higher than the threshold value as seed points, and triggering a growth process in sequence according to the order of scores from high to low; the growth cost assessment, namely initiating growth detection to adjacent grids by taking seed points as centers, and calculating the growth cost of merging the adjacent grids into the current land, wherein the growth cost is formed by weighting attribute cost reflecting the quality difference of cultivated land and topology cost reflecting the space resistance; and when the growth cost is smaller than a preset condition, merging the adjacent grids into the current land block, and updating the land block boundary.
6. The permanent basic farmland intelligent repayment management method based on data processing according to claim 5, wherein the topology cost is calculated based on anisotropic terrain distance, The calculation process of the anisotropic terrain distance merges the terrain gradient direction constraint, and the specific logic is as follows: calculating the included angle between the current growth direction vector and the topographic gradient direction vector of the position in real time, When the growth direction is perpendicular to the direction of the terrain gradient, a first penalty factor is given, When the growth direction is parallel to the direction of the topographical gradient, a second penalty factor is assigned, wherein, The second penalty factor is significantly greater than the first penalty factor.
7. The permanent basic farmland intelligent repair and scratch management method based on data processing according to claim 1, wherein in S400, the process of constructing a multi-objective optimization function includes: defining decision variables, namely distributing a binary decision variable for each preliminary continuous land block generated in the step S300; calculating an area weighted average suitability score of the selected land block combination as a maximization target; Constructing a space breaking degree punishment subfunction, namely calculating the ratio of the perimeter of the boundary to the square root of the area of each selected land as a shape index, and taking the accumulated value of the shape indexes as a minimization target; and carrying out global solution on the decision variables by utilizing a genetic algorithm, and searching for a pareto optimal solution on the premise that the sum of the areas of the selected land pareto is larger than a preset target value.
8. The permanent basic farmland intelligent repair and scratch management method based on data processing according to claim 1, characterized in that the intelligent repair and scratch management method further comprises: S500, based on the land block boundary determined by the optimal complement and division scheme, historical time sequence remote sensing image data are called to carry out data backtracking verification, abnormal land blocks with non-agricultural construction traces are removed, and final complement and division land block data are output.
9. The method for intelligent repair and scratch management of permanent basic farmland based on data processing according to claim 8 is characterized in that the method for retrieving historical time sequence remote sensing image data for data backtracking verification comprises the steps of extracting time sequence data of normalized vegetation index NDVI of a target land block in a historical preset time period, calculating the mean value, variance and change rate of the NDVI time sequence data to construct a vegetation fluctuation abnormal index, and judging that a historical suspicious point exists in the land block and generating an early warning signal if the vegetation coverage change rate of the land block exceeds a preset natural replacement threshold or the NDVI curve characteristic of the land block shows a mathematical mode of non-agricultural construction trace.
10. The utility model provides a permanent basic farmland intelligence benefit is drawn management system based on data processing which characterized in that, intelligent benefit is drawn management system includes: A processor; a memory storing a computer program which, when executed by a processor, implements the permanent basic farmland intelligent repair and scratch management method based on data processing according to any one of claims 1 to 9.

Description

Permanent basic farmland intelligent repair and scratch management method and system based on data processing Technical Field The invention relates to the field of geographic information systems and homeland resource space planning, in particular to a permanent basic farmland intelligent repair and planning management method and system based on data processing. Background The permanent basic farmland is a red line of national grain safety, and along with the promotion of the urbanization process, part of the original permanent basic farmland is inevitably occupied, so that the equal quality and equal quantity of repair and scratch are required to be carried out in other areas. At the present stage, the traditional repair work faces a great technical challenge: Firstly, the prior art only carries out simple deduction and screening according to a single land utilization current situation map spot, ignores comprehensive influences of multidimensional factors such as terrain gradient, soil physicochemical property, surrounding ecological environment and the like, and leads to that a defined farmland is nominally cultivated, but the actual cultivation condition is poor, the farmland is difficult to be used for a long time and related regulations of a permanent basic farmland cannot be met. Secondly, the traditional manual screening or simple GIS superposition analysis is easy to generate a large number of scattered and finely-crushed spot-shaped plots which are distributed discretely and severely crushed in space, is unfavorable for large-scale operation, cannot meet the requirements of modern agricultural mechanized operation, and increases the management cost. And when the hilly and mountain areas in the southwest area are complemented, if only the plane distance is considered and the directivity of the terrain gradient is ignored, the land block which is cultivated along the slope is easily marked, so that the cultivation difficulty is increased, and serious water and soil loss is easily caused. Finally, in order to complete the index in a part of areas, the land block with the building just removed, the soil not restored and even only the surface soil paved is possibly brought into the repair and scratch range, and the historical non-agricultural construction trace is difficult to identify only by the image data at the current moment, so that the repair and scratch result has the risks of high quality and poor stability. Therefore, there is a need for an intelligent complement management method that can integrate multi-source data, compromise spatial morphology optimization, adapt to complex terrain constraints, and have historical authenticity verification capability. Disclosure of Invention The invention aims to provide a permanent basic farmland intelligent compensation management method based on data processing, which aims to solve the problems of uneven quality, broken space layout, poor terrain adaptability and lack of historical authenticity supervision of a compensation land block in the prior art. The permanent basic farmland intelligent patch management method based on data processing comprises the following steps of S100, obtaining a multi-source geospatial data set of a target area, mapping the multi-source geospatial data set into a unified grid system, constructing a multi-dimensional feature vector of each grid unit, S200, calculating a cultivated land suitability score of each grid unit by using a cultivated land suitability evaluation model comprising a space smoothness constraint mechanism based on the multi-dimensional feature vector, generating a cultivated land suitability distribution map, S300, based on the cultivated land suitability distribution map, utilizing an anisotropic region growing algorithm to aggregate discrete grid units into a plurality of preliminary patch plots, wherein the anisotropic region growing algorithm fuses a topographic gradient direction constraint in a growing process, S400, constructing a multi-objective optimization function comprising a cultivated land quality maximization target and a space breakage minimization target by taking the patch area as a rigid constraint, and screening the optimal patch scheme from the preliminary patch plots. Further, the multisource geospatial data set refers to heterogeneous data combinations for describing natural geographic attributes, agricultural production conditions and space control states of a target area in an omnibearing and multi-granularity mode, and the data set comprises, but is not limited to, high-resolution remote sensing image data for extracting vegetation coverage features of the ground surface, digital elevation model data for constructing three-dimensional terrain features, soil physicochemical property data for evaluating the soil physical and chemical properties of cultivated land force, and utilization status and planning vector data for legal attributes and space boundaries of a fixed land block. F