Land contractor change identification method and system based on remote sensing image

Abstract

A land contractor change identification method and system based on a remote sensing image relates to the technical field of agricultural remote sensing monitoring, and the method comprises the following steps: calculating vegetation index time sequence data of the remote sensing image, determining a vegetation index mutation time point as farming activity time based on a time dimension mean value change rate of the vegetation index time sequence data; and calculating the standard deviation mean value of the vegetation index time sequence data in each period in the spatial dimension as the tillage uniformity. And calculating a time difference value and a uniformity difference value of adjacent land parcels, performing normalization processing on the two types of difference values, and performing weighting according to a preset weight to obtain a farming activity difference index. And when the difference index of the target land parcel is greater than a preset difference threshold value, marking the target land parcel as a suspected contract change land parcel. And when the correlation coefficient of the difference index time sequence curve of the suspected changed land parcel and the historical time sequence curve is smaller than a preset correlation threshold value, determining the suspected changed land parcel as a contractor changed land parcel. By implementing the method, the identification accuracy of the land contractor change can be improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of agricultural remote sensing monitoring, and particularly relates to a land contract right change identification method and system based on remote sensing images. BACKGROUND

[0002] With the deepening of the rural land contract management right system, it is of great significance to maintain the stability of rural land rights and interests to discover and monitor the land contract right change in a timely manner. The traditional field investigation method is time-consuming and labor-intensive, and the remote sensing technology has the advantages of wide range and high frequency of monitoring, and has become an important means of land change monitoring.

[0003] The existing remote sensing change detection technology mainly analyzes the differences in surface coverage features. This technology extracts spectral, texture and other features in remote sensing images, calculates the change amount between images at different times, and judges whether the surface has changed. In specific implementation, first, the features of the remote sensing image are extracted, and then the position and range of the change area are determined by setting a change threshold.

[0004] However, in the farmland monitoring scene, the change of the land plot is often affected by the growth cycle of crops. Due to seasonal farming activities, the periodic change of the surface features makes it difficult to distinguish the short-term change caused by farming activities from the long-term change rule caused by the change of the land contract right, and misjudgment is easy to occur. SUMMARY

[0005] The present application provides a land contract right change identification method and system based on remote sensing images, which is used to improve the identification accuracy of the land contract right change.

[0006] In a first aspect, the application provides a land contract right change identification method based on remote sensing images, applied to a land contract right change identification system. The method comprises: acquiring continuous multi-period high-resolution remote sensing images in a target research area, the high-resolution remote sensing images being continuously acquired at a set time interval, and the high-resolution remote sensing images including images in a key farming activity period in a crop growing season; calculating vegetation index time series data using the high-resolution remote sensing images; determining a vegetation index mutation time point as a farming activity time according to a mean rate of change of the vegetation index time series data in the time dimension, and calculating a mean standard deviation of the vegetation index time series data in the spatial dimension as a tillage uniformity, the vegetation index mutation time point being a time point at which the change rate of adjacent two periods of vegetation index time series data is greater than a preset change rate threshold; obtaining a time difference value according to the absolute difference of the farming activity time of adjacent plots, obtaining a uniformity difference value according to the absolute difference of the tillage uniformity of adjacent plots, and obtaining a farming activity difference index by weighting the time difference value and the uniformity difference value after normalization processing according to a preset weight; when the farming activity difference index of a target plot is greater than a preset difference threshold, marking the target plot as a suspected contract right change plot; and when the correlation coefficient of the farming activity difference index time series curve of the suspected contract right change plot and a historical time series curve is less than a preset correlation threshold, determining the suspected contract right change plot as a contract right change plot.

[0007] In the above embodiment, the high-resolution remote sensing image time series analysis is combined with the farming activity synchronicity evaluation, the farming activity time and the tillage uniformity are calculated based on the vegetation index time series data, the difference index reflecting the coordination degree of the farming activities of adjacent plots is constructed, the correlation analysis of the historical time series pattern is performed, the long-term change rule caused by the contract right change and the short-term change caused by the farming activity are effectively distinguished, the real contract right change plot is accurately identified, and the problem of misjudgment of normal farming activity difference as contract right change is effectively avoided.

[0008] In combination with some embodiments of the first aspect, in some embodiments, the step of determining the vegetation index mutation time point as the farming activity time according to the mean rate of change of the vegetation index time series data in the time dimension specifically comprises: dividing the vegetation index time series data into a plurality of observation periods according to a set observation period; calculating the daily mean value of the vegetation index in each observation period; calculating the difference between the daily mean values of the vegetation index of adjacent two observation periods to obtain a vegetation index change amount, and calculating the mean rate of change of the vegetation index time series data according to the vegetation index change amount; extracting the time point at which the mean rate of change is greater than a preset change rate threshold as the vegetation index mutation time point, and determining the farming activity time according to the distribution characteristics of the vegetation index mutation time point, the farming activity time being a time point in a time series set at which the density value is greater than a preset density threshold.

[0009] In the above embodiment, the vegetation index time series data is divided according to the observation period and the daily average is calculated, the change amount and the change rate of adjacent periods are analyzed, the distribution density characteristics of the vegetation index mutation time points are extracted, the multi-level recognition mechanism of the farming activity time is established, the accurate positioning of the farming activity time is realized, a reliable time benchmark is provided for the subsequent synchronization analysis, and the time precision and the reliability of the contract right change recognition are improved.

[0010] In combination with some embodiments of the first aspect, in some embodiments, after the time difference value and the uniformity difference value are normalized, the step of obtaining the farming activity difference index according to the preset weight, specifically includes: calculating the time span of the farming activity time of each plot in the research area, and dividing the time difference value by the time span to obtain the normalized time difference value; calculating the change range of the tillage uniformity of each plot in the research area, and dividing the uniformity difference value by the change range to obtain the normalized uniformity difference value; determining the time difference weight and the uniformity difference weight based on the indication degree of the farming activity to the change of the land contract right, and multiplying the normalized time difference value by the time difference weight, the normalized uniformity difference value by the uniformity difference weight, and then adding them to obtain the farming activity difference index.

[0011] In the above embodiment, the span of the farming activity time of the plot in the research area and the change range of the tillage uniformity are normalized, the weight coefficient is determined in combination with the indication degree of the farming activity to the change of the land contract right, a comprehensive farming activity difference index is constructed, the unified quantitative evaluation of different types of farming activities is realized, and the contract right change recognition result has stronger comparability and consistency.

[0012] In combination with some embodiments of the first aspect, in some embodiments, when the correlation coefficient of the farming activity difference index time series curve of the suspected contract right change plot and the historical time series curve is less than the preset correlation threshold, the step of determining the suspected contract right change plot as the contract right change plot, specifically includes: obtaining the farming activity difference index time series data of the suspected contract right change plot in a plurality of historical years without contract right change; performing average processing on the farming activity difference index time series data of the plurality of historical years to obtain a baseline historical time series curve; collecting the farming activity difference index of each farming activity period in the current year of the suspected contract right change plot to generate a current time series curve; calculating the correlation coefficient of the current time series curve and the baseline historical time series curve; when the correlation coefficient is less than the preset correlation threshold, the suspected contract right change plot is determined as the contract right change plot.

[0013] In the above embodiment, the difference index data of the suspected contract right change plot in multiple historical years is averaged to generate a standard baseline historical time series curve, the change is determined by calculating the correlation coefficient of the current time series curve and the baseline curve, a dynamic comparison verification mechanism based on historical data is established, the accuracy and reliability of the contract right change recognition are significantly improved, and the misjudgment influence caused by environmental factors and random fluctuations is reduced.

[0014] In combination with some embodiments of the first aspect, in some embodiments, after the step of determining the suspected contract right change plot as a contract right change plot when the correlation coefficient of the agricultural activity difference index time series curve of the suspected contract right change plot and the historical time series curve is less than the preset threshold, the method further comprises: obtaining the agricultural activity difference index time series curve of the surrounding plots of the contract right change plot; dividing the surrounding plots into a synchronous change area and a non-synchronous change area according to the similarity of the agricultural activity difference index time series curves of the contract right change plot and the surrounding plots; calculating the spatial autocorrelation coefficient of the agricultural activity difference index of the plots in the synchronous change area, and determining the concentrated and contiguous area of the contract right change according to the spatial autocorrelation coefficient.

[0015] In the above embodiment, the difference index time series curve of the surrounding area of the contract right change plot is obtained, the area is divided based on the similarity of the time series characteristics, the spatial autocorrelation coefficient of the plots in the synchronous change area is calculated, and a spatial contiguous recognition mechanism of the contract right change is established, thereby realizing the automatic delineation and range determination of the concentrated and contiguous change area, and effectively revealing the spatial aggregation rule and diffusion trend of the contract right change.

[0016] In combination with some embodiments of the first aspect, in some embodiments, the step of dividing the surrounding plots into a synchronous change area and a non-synchronous change area according to the similarity of the agricultural activity difference index time series curves of the contract right change plot and the surrounding plots specifically comprises: establishing an agricultural activity difference index time series feature vector of the contract right change plot, the time series feature vector including fluctuation amplitude, periodicity and trend; calculating the similarity coefficient between the agricultural activity difference index time series feature vector of the surrounding plot and the time series feature vector of the contract right change plot; determining a time series feature similarity score according to the similarity coefficient, the time series feature similarity score reflecting the consistency degree of the agricultural activity change; and dividing the surrounding plots with the time series feature similarity score greater than a set threshold into the synchronous change area, and dividing the surrounding plots with the time series feature similarity score less than or equal to the set threshold into the non-synchronous change area.

[0017] In the above embodiment, the time sequence feature vector containing fluctuation amplitude, periodicity and trend is constructed, the feature vector similarity coefficient of the surrounding land and the change land is calculated, the region is classified based on the similarity score, the time sequence pattern matching mechanism of multi-dimensional features is established, the fine identification and differentiation of different change types are realized, and the accuracy and reliability of the contract right change region division are improved.

[0018] In combination with some embodiments of the first aspect, in some embodiments, after the step of determining the suspected contract right change land as the contract right change land when the correlation coefficient of the time sequence curve of the agricultural activity difference index of the suspected contract right change land and the historical time sequence curve is less than the preset threshold, the method further includes: obtaining the field verification result of the contract right change land in the historical period, and establishing a corresponding relationship between the time sequence curve of the agricultural activity difference index and the field verification result; identifying the agricultural activity feature mode of different types of land contract right change according to the corresponding relationship, and determining the change type of the newly identified contract right change land based on the agricultural activity feature mode.

[0019] In the above embodiment, the historical verification result is obtained and the corresponding relationship with the time sequence curve is established, the agricultural activity feature mode of different types of change is identified, the change type discrimination standard based on historical experience is formed, the adaptive change mode learning mechanism is constructed, the intelligent judgment of the type of the newly identified change land is realized, the automation level and adaptability of the contract right change type identification are effectively improved, and more accurate change information support is provided for subsequent management decision.

[0020] In the second aspect, the embodiments of the present application provide a land contract right change identification system, which comprises one or more processors and a memory; the memory is coupled with the one or more processors, the memory is used to store computer program code, the computer program code comprises computer instructions, and the one or more processors invoke the computer instructions to make the land contract right change identification system execute the method described in the first aspect and any possible implementation manner of the first aspect.

[0021] In the third aspect, the embodiments of the present application provide a computer program product containing instructions, when the above computer program product runs on the land contract right change identification system, the above land contract right change identification system executes the method described in the first aspect and any possible implementation manner of the first aspect.

[0022] In the fourth aspect, the embodiments of the present application provide a computer readable storage medium, which comprises instructions, when the above instructions run on the land contract right change identification system, the above land contract right change identification system executes the method described in the first aspect and any possible implementation manner of the first aspect.

[0023] It can be understood that the land contract right change identification system provided in the second aspect, the computer program product provided in the third aspect and the computer storage medium provided in the fourth aspect are all used to execute the method provided in the embodiments of the present application. Therefore, the beneficial effects that can be achieved thereby can refer to the beneficial effects in the corresponding method, which will not be described here again.

[0024] The one or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages: 1. In the present application, high-resolution remote sensing images are combined with the evaluation of the synchronism of agricultural activities, the time of agricultural activities and the uniformity of cultivation are calculated based on the time series data of vegetation index, the difference index reflecting the coordination degree of agricultural activities of adjacent plots is constructed, the correlation analysis of historical time series patterns is performed, the long-term change rule caused by the change of contract right and the short-term change caused by agricultural activities are effectively distinguished, the real contract right change plot is accurately identified, and the problem of misjudgment of normal agricultural activity difference as contract right change is effectively avoided.

[0025] 2. In the present application, the time series data of vegetation index is divided according to the observation period and the daily average is calculated, the change amount and the change rate of adjacent periods are analyzed, the distribution density characteristics of the time point of vegetation index mutation are extracted, the multi-level identification mechanism of the time of agricultural activities is established, the accurate positioning of the time of agricultural activities is realized, a reliable time reference is provided for the subsequent synchronism analysis, and the time accuracy and reliability of the identification of contract right change are improved.

[0026] 3. In the present application, the difference index data of the suspected contract right change plot in multiple historical years is extracted for average processing to generate a standard reference historical time series curve, the correlation coefficient between the current time series curve and the reference curve is calculated for change determination, a dynamic comparison and verification mechanism based on historical data is established, the accuracy and reliability of the identification of contract right change are significantly improved, and the misjudgment influence caused by environmental factors and random fluctuations is reduced. BRIEF DESCRIPTION OF DRAWINGS

[0027] Figure 1 is a flowchart of a land contract right change identification method based on remote sensing images in the embodiments of the present application; Figure 2 is another flowchart of a land contract right change identification method based on remote sensing images in the embodiments of the present application; Figure 3 is a schematic diagram of an entity device structure of a land contract right change identification system in the embodiments of the present application. DETAILED DESCRIPTION

[0028] The terminology used in the following embodiments of the present application is for the purpose of describing particular embodiments only and is not intended to be limiting of the present application. As used in the description of the application, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms "and / or", as used herein, refers to

[0029] Hereinafter, the terms "first", "second" are only for the purpose of description, and cannot be understood as implying or suggesting relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined with "first", "second" can explicitly or implicitly include one or more of the features, and in the description of the embodiments of the present application, the meaning of "multiple" is two or more, unless otherwise specified.

[0030] For the convenience of understanding, the application scenarios of the embodiments of the present application are introduced as follows.

[0031] In the rural land contract management right confirmation registration work of a certain province, the task of monitoring the change of land contract right in a large range of farmland is faced. The traditional manual field investigation method needs to invest a lot of manpower and material resources, the investigation period is long, and the investigation frequency is low, so it is difficult to find the change of contract right in time. With the deepening of the rural land circulation policy, the behaviors of land subletting, leasing and exchanging among farmers are becoming more and more frequent, and it is urgent to establish an efficient monitoring system to maintain the stability of rural land rights.

[0032] Remote sensing technology provides a new technical path to solve this problem with its advantages of wide range and high frequency monitoring. By obtaining continuous multi-period high-resolution remote sensing images in the study area, dynamic monitoring of farmland area can be realized. However, the particularity of farmland environment is that the change of land is often affected by the growth cycle of crops, and seasonal farming activities will cause periodic changes in surface features. Sowing, fertilizing, harvesting and other farming activities will produce obvious spectral and texture changes in remote sensing images. These normal agricultural production activities and real land contract right changes may show similar image features.

[0033] How to accurately identify the change pattern caused by real land contract right change from these complex time series changes and avoid misjudging normal farming activities as contract right change becomes a key challenge for technical implementation. Under this background, a technical method is needed to distinguish between short-term changes caused by farming activities and long-term change rules caused by contract right changes.

[0034] The existing remote sensing change detection technology mainly analyzes the differences in surface cover features, which has obvious technical limitations in farmland monitoring applications. This technology extracts spectral, texture and other features from remote sensing images, calculates the change between different periods of images, and determines the location and range of the change area by setting a change threshold.

[0035] In practical applications, this method faces significant challenges. During spring plowing, adjacent plots show different surface features in the same period due to differences in planting times. Even adjacent plots managed by the same contractor, due to different crop varieties or slight differences in planting times, their vegetation index time series curves will also differ significantly, and the traditional method is prone to misjudgment of normal farming activity differences as land contract right changes.

[0036] The situation at the harvest period is also complex, and the differences in harvesting times of different plots will cause the vegetation coverage to change significantly over time. The traditional method has difficulty distinguishing between short-term farming activity changes and real contract right changes, and often identifies normal harvest time differences as contract right change signals.

[0037] More importantly, when a real contract right change occurs, the new contractor will often continue the original farming mode and continue to plant the same or similar crops, and adopt similar farming activity time arrangements. In this case, the surface feature changes are not obvious, and the traditional method based on spectral differences is prone to miss such real contract right changes.

[0038] The improved method of the present application achieves more accurate identification through farming activity synchronicity analysis. The scheme first acquires continuous multi-period high-resolution remote sensing images in the study area, ensuring complete coverage of the crop growing season, including images of key farming activity periods such as sowing, growing, maturity, and harvesting.

[0039] The system can accurately identify the vegetation index mutation time point as the farming activity time by calculating the vegetation index time series data. When the change rate of the vegetation index time series data of adjacent two periods is greater than the preset change rate threshold, the system marks the time point as the vegetation index mutation time point, and then determines the specific time of the farming activity. At the same time, the system calculates the standard deviation mean of each period of vegetation index time series data in the spatial dimension as the farming uniformity index, which is used to quantify the spatial consistency characteristics of farming activities.

[0040] The core innovation lies in the construction and application of the agricultural activity difference index. The system calculates the absolute difference value of the agricultural activity time of adjacent plots to obtain the time difference value, calculates the absolute difference value of the uniformity of adjacent plots to obtain the uniformity difference value, and after normalization processing of the two difference values, the comprehensive agricultural activity difference index is obtained by weighting according to the preset weight. This index can effectively reflect the synchronization degree of the agricultural activities of adjacent plots.

[0041] When the agricultural activity difference index of the target plot is greater than the preset difference threshold, the system marks it as a suspected contract right change plot. Further, the system compares the correlation coefficient of the agricultural activity difference index time series curve of the suspected plot with the historical time series curve, and when the correlation coefficient is less than the preset correlation threshold, it is confirmed as a real contract right change. This multi-level verification mechanism effectively avoids the problem of misjudging normal agricultural activity differences as contract right changes.

[0042] In addition, the system also has the ability to identify the concentrated contiguous area of contract right change, and determines the spatial distribution range of the contract right change by analyzing the similarity of the agricultural activity difference index time series curve of the surrounding plots and performing spatial autocorrelation analysis. Combined with the feature mode library established based on historical field verification data, the system can also automatically distinguish different types of contract right change modes.

[0043] For ease of understanding, the method provided by the present embodiment will be described in the following flowchart in conjunction with the above scenario. Please refer to Figure 1 , a flowchart of the method for identifying land contract right change based on remote sensing images in the embodiments of the present application.

[0044] S101, acquire continuous multi-period high-resolution remote sensing images in a target research area, the high-resolution remote sensing images are acquired continuously according to a set time interval, and the high-resolution remote sensing images include images of key agricultural activity periods in the crop growing season.

[0045] Among them, the target research area represents a specific geographical range that needs to be monitored for land contract right change, which is usually a concentrated farmland distribution area delineated based on administrative boundaries or natural geographical boundaries. Continuous multi-period high-resolution remote sensing images refer to high spatial resolution satellite image data acquired according to a fixed time sequence, which can clearly identify plot boundaries and surface cover characteristics. The set time interval is used to represent the frequency of image acquisition, which is usually determined according to the crop growth cycle and the characteristics of agricultural activities. The key agricultural activity period refers to important time nodes in the crop growing season that have a significant impact on changes in ground features, including sowing period, seedling period, jointing period, heading period, filling period, maturity period and harvesting period. High-resolution remote sensing images can come from various satellite platforms, such as Sentinel series, Landsat series, high-resolution (GF) series satellites, or obtained by unmanned aerial photography, which are not limited in the present application.

[0046] Specifically, the system first determines the geographic range and coordinate boundary of the target research area according to the land contract right change monitoring requirements, and then formulates a detailed image acquisition plan. The image acquisition needs to cover the entire crop growing season to ensure that the surface change characteristics from spring planting to autumn harvest can be captured. The system will determine the appropriate image acquisition time interval according to the agricultural production characteristics and crop planting structure of the research area, usually once every 7-15 days, to ensure that key changes in agricultural activities can be captured while avoiding data redundancy. In terms of image quality control, the system will prioritize high-quality images with a cloud cover of less than 10% to ensure the clarity and accuracy of the surface information. At the same time, the system will establish a spatial registration and radiation correction process for image data to eliminate geometric deviations and radiation differences between images of different periods, laying a reliable data foundation for subsequent time series analysis.

[0047] In some embodiments, the acquisition and preprocessing of high-resolution remote sensing images can be achieved in various ways: optionally, through the satellite data service platform to automatically order and download remote sensing image data of the target area, the system will automatically select images that meet the requirements according to the preset time plan and quality standards, then perform batch download and storage management, establish a complete image database, and record and evaluate the quality of each image metadata to ensure data integrity and traceability; optionally, a multi-source remote sensing data fusion mechanism is established to simultaneously acquire high-resolution image data from different satellite platforms, improve the temporal resolution and spatial coverage through data fusion technology, and the system will perform unified geometric correction and radiation normalization processing on image data from different sources, then select and combine according to image quality and time matching degree to form a high-quality continuous time series image sequence. It can be understood that other ways of acquiring and managing remote sensing image data can also be used, which are not limited here.

[0048] S102, calculate the vegetation index time series data using high-resolution remote sensing images.

[0049] Among them, the vegetation index represents a numerical index calculated from the multispectral band information of the remote sensing image to quantify the vegetation coverage and growth status. Commonly used vegetation indices include normalized difference vegetation index NDVI, enhanced vegetation index EVI, soil-adjusted vegetation index SAVI, etc. Time series data refers to a sequence of vegetation index values arranged in chronological order, which can reflect the dynamic change characteristics of vegetation coverage over time. The calculation process is used to represent the technical process of extracting vegetation information from the spectral reflectance data of the original remote sensing image and generating standardized vegetation indices. The vegetation index time series data is used to represent the vegetation index change trajectory of each plot or pixel during the entire research period, which is the core data basis for subsequent agricultural activity recognition and contract right change analysis.

[0050] Specifically, the system first performs radiation correction and atmospheric correction on the acquired high-resolution remote sensing images, converts the original digital quantization values into surface reflectance data, and eliminates the effects of atmospheric scattering and absorption on spectral information. Then, the system calculates various vegetation indices according to the calculation formulas of different vegetation indices using the corresponding spectral band information. Taking NDVI as an example, the system uses the reflectance data of the near-infrared band and the red band to calculate the NDVI value of each pixel according to the formula NDVI = (NIR-Red) / (NIR+Red), where NIR represents the near-infrared band reflectance and Red represents the red band reflectance. The system calculates the vegetation index for all pixels in each image in the study area to generate the vegetation index image for the corresponding period. Subsequently, the system extracts the vegetation index statistical values of each plot according to the plot boundaries, including mean, median, standard deviation, and other statistical quantities, to construct the vegetation index time series data matrix of each plot. To ensure data quality, the system also identifies and processes outliers, uses a time series smoothing algorithm to eliminate data noise caused by factors such as cloud shadows and haze, and generates continuous and reliable vegetation index time series curves.

[0051] In some embodiments, the calculation of vegetation indices and the construction of time series data can be achieved in various ways: optionally, a multi-vegetation index fusion strategy can be used to improve the accuracy of vegetation information extraction. The system calculates multiple vegetation indices such as NDVI, EVI, and SAVI, then performs weighted fusion according to the characteristics and application conditions of different indices to generate a comprehensive vegetation index, reduces the limitations of a single index through the complementarity of multiple index information, improves the sensitivity to different crop types and growth stages, and ultimately forms more stable and reliable vegetation index time series data; optionally, a vegetation index optimization algorithm based on machine learning can be established. The system analyzes the relationship between historical vegetation index data and ground measurement data, trains a vegetation index optimization model, automatically adjusts the vegetation index calculation parameters and weight coefficients, improves the adaptability of vegetation indices to specific research areas, and simultaneously uses deep learning technology to identify and repair outliers and missing values in time series data to generate a high-quality continuous vegetation index time series data set. It can be understood that other ways can also be used to realize the calculation of vegetation indices and the generation of time series data, which are not limited here.

[0052] Commonly used vegetation indices include normalized difference vegetation index NDVI, enhanced vegetation index EVI, and soil-adjusted vegetation index SAVI. Those skilled in the art can understand that combinations of these indices or other indices that can reflect vegetation growth conditions can also be used, and the present application is not limited.

[0053] S103, determining a vegetation index mutation time point as a farming activity time according to a mean rate of change of the vegetation index time series data in the time dimension, and calculating a mean of standard deviations of each period of vegetation index time series data in the spatial dimension as a tillage uniformity, the vegetation index mutation time point being a time point at which a change rate of adjacent two periods of vegetation index time series data is greater than a preset change rate threshold.

[0054] The mean rate of change in the time dimension represents the average change amplitude of the vegetation index between adjacent time nodes, and the change rate of the vegetation coverage is quantified by calculating the ratio of the difference between the mean values of the adjacent two periods of vegetation index to the time interval. The vegetation index mutation time point refers to a specific time at which the vegetation index has a significant jump change, and these time points usually correspond to the time of important farming activities, such as sowing, fertilizing, harvesting, and other key agricultural operations. The farming activity time is used to represent the actual occurrence time of each specific operation in the agricultural production process, and is an important time marker for judging the management mode and change of the contract right of the land plot. The mean of standard deviations in the spatial dimension represents a statistical quantity of the dispersion degree of the vegetation index at different positions within the land plot in the same period, reflecting the spatial uniformity characteristics of the vegetation coverage within the land plot. The tillage uniformity refers to the consistency degree of the spatial distribution of the farming activities, and a high tillage uniformity indicates that the management of the land plot is uniform, and a low tillage uniformity may indicate a change in the management mode. The preset change rate threshold is used to represent the critical standard for determining whether the vegetation index has a mutation, and is usually determined according to the experience of the agricultural production characteristics and crop types of the study area.

[0055] Specifically, the system first analyzes the change rate in the time dimension of the vegetation index time series data of each land plot, calculates the difference between the mean values of the vegetation index of adjacent two observation periods, and divides the time interval to obtain a standardized change rate value. The system will set a reasonable change rate threshold, and when the absolute value of the calculated change rate exceeds the threshold, the corresponding time point will be marked as a vegetation index mutation time point. In order to accurately identify the farming activity time, the system will perform clustering analysis on the identified mutation time points, and merge the mutation points close in time into the same farming activity time window, so as to obtain the occurrence time of the key farming activities of each land plot. In terms of tillage uniformity calculation, the system will extract all pixel vegetation index values of the land plot in each observation period, calculate the standard deviation of these values, and then average all standard deviation values in the study period to obtain the tillage uniformity index of the land plot. The system also establishes a dynamic threshold adjustment mechanism to adaptively adjust the change rate threshold according to the growth characteristics of different seasons and crop types, improving the accuracy and adaptability of the identification of the farming activity time.

[0056] In some embodiments, the vegetation index mutation point detection and the plowing uniformity calculation can be implemented in various ways. Optionally, a multi-scale mutation detection method based on wavelet transform is adopted. The system first performs wavelet decomposition on the vegetation index time series data to obtain the change characteristics at different time scales, then calculates the change rate at each scale and sets the corresponding threshold standard, improves the robustness of mutation point detection through multi-scale information fusion, identifies the accurate time position of the mutation point using continuous wavelet transform, verifies the correspondence between the mutation point and the actual farming activities in combination with the agricultural production calendar, and finally generates a reliable farming activity time sequence. Optionally, an adaptive threshold determination mechanism based on statistical learning is established. The system analyzes the correspondence between the historical vegetation index change pattern and the known farming activity time, trains the threshold optimization model, automatically learns the optimal change rate threshold for different plots, different crop types and different growth stages, uses the Bayesian method to process the uncertainty of threshold determination, continuously improves the accuracy of mutation point detection through iterative optimization, and establishes a confidence evaluation mechanism to control the quality of the detection results. It can be understood that other methods can also be used to implement vegetation index mutation detection and farming activity time determination, which are not limited here.

[0057] In some embodiments, the step specifically includes the following steps: The vegetation index time series data is divided into multiple observation periods according to the set observation period, the daily average of the vegetation index in each observation period is calculated, the difference between the daily average of the vegetation index of adjacent observation periods is calculated to obtain the vegetation index change amount, the mean change rate of the vegetation index time series data is calculated according to the vegetation index change amount, the time points with a mean change rate greater than a preset change rate threshold are extracted as vegetation index mutation time points, and the farming activity time is determined according to the distribution characteristics of the vegetation index mutation time points. The farming activity time is the time point with a density value greater than a preset density threshold in the time sequence set located at the vegetation index mutation time point.

[0058] The vegetation index time series data is a sequence of vegetation coverage values arranged in chronological order calculated from remote sensing images. The set observation period is a time interval determined according to the characteristics of farming activities, usually set to 7-15 days. The observation period refers to the continuous time interval divided according to the observation period. The daily average of the vegetation index is the arithmetic mean of the vegetation index on all dates in the observation period. The vegetation index change amount is the numerical difference between the daily average of adjacent observation periods. The mean change rate is the ratio of the change amount to the time interval, reflecting the change speed of the vegetation index. The vegetation index mutation time point is a specific time when the change rate exceeds the threshold, corresponding to the occurrence time of important farming activities. The distribution characteristics refer to the aggregation and dispersion rules of the mutation time points on the time axis. The density value is a quantitative indicator of the concentration of mutation time points within a specific time window. The farming activity time is the time when the key agricultural operation is confirmed through density screening.

[0059] When performing this step, first, the complete vegetation index time series data T = [NDVI1, NDVI2,..., NDVIn] is time divided according to the set observation period Δt, generating consecutive observation periods Ti = [ti_start, ti_end], where i is the period number, ti_start and ti_end are the start and end times of the i-th period, respectively. The daily mean vegetation index in each observation period is calculated: NDVIi_mean = Σ(NDVIj) / ni, where j is the date index within the period Ti, and ni is the number of days within the period Ti. The daily mean time series M = [NDVI1_mean, NDVI2_mean,..., NDVIm_mean] is constructed, where m is the total number of observation periods. The vegetation index change amount of adjacent observation periods is calculated: ΔNDVIi = NDVIi+1_mean - NDVIi_mean, obtaining the change amount sequence Δ = [ΔNDVI1, ΔNDVI2,..., ΔNDVIm-1]. The mean change rate is calculated according to the change amount and time interval: Ratei = ΔNDVIi / Δt, forming the change rate time series R = [Rate1, Rate2,..., Ratem-1]. A preset change rate threshold Rthreshold is set, usually determined based on historical data statistics, with a value range of 0.02-0.05 / day. The change rate time series is traversed, and when |Ratei| > Rthreshold, the corresponding time point ti is marked as a vegetation index mutation time point, and the mutation time point set C = {t1, t2,..., tk} is constructed, where k is the total number of mutation points. The distribution characteristics of the mutation time points are analyzed using the kernel density estimation method, and a time window width w is set (usually 10-15 days), and the density value at each time point τ is calculated: Density(τ) = (1 / k) × Σ K((τ-ti) / w), where K is the Gaussian kernel function. A preset density threshold Dthreshold is set, usually taking the 80th percentile value of the density distribution. The time regions with density values greater than the threshold are identified, and the time points corresponding to the density peak values in each region are extracted as the agricultural activity time: A = {τ | Density(τ) > Dthreshold and is a local maximum}. The correspondence between the identified agricultural activity time and the actual agricultural production calendar is verified, and obviously abnormal time points are removed, generating the final agricultural activity time sequence, providing an accurate time reference for subsequent synchronization analysis.

[0060] S104, obtain a time difference value according to an absolute difference value of farming activity time of adjacent plots, obtain a uniformity difference value according to an absolute difference value of farming uniformity of adjacent plots, normalize the time difference value and the uniformity difference value, and then obtain a farming activity difference index according to a preset weight.

[0061] The adjacent plots represent farmland plots directly adjacent or close in spatial position, usually have similar natural conditions and agricultural production environment, and should present similar farming activity patterns under normal circumstances. The time difference value refers to the absolute difference degree of farming activity time between adjacent plots, and the greater the value, the worse the time synchronization of farming activity of adjacent plots. The uniformity difference value is used to represent the absolute difference between the farming uniformity indexes of adjacent plots, and reflects the difference in the fine management degree of adjacent plots. The normalization processing is to convert the difference values of different dimensions and value ranges into values in a unified standardized interval, eliminate the dimension influence and make different types of difference values comparable. The preset weight represents the relative importance coefficient of time difference and uniformity difference in the evaluation of farming activity synchronization, which is determined according to agricultural expert knowledge and historical data analysis. The farming activity difference index is used to represent the comprehensive quantitative index of the coordination degree of adjacent plots, and high synchronization usually indicates that the plots are operated by the same management subject, and low synchronization may indicate that the contract right is changed.

[0062] Specifically, the system first establishes a topological structure of plot spatial adjacency relationship, identifies the adjacent plot set of each plot, and then calculates the farming activity time difference between the target plot and its adjacent plots one by one. For each pair of adjacent plots, the system extracts their respective key farming activity time series, calculates the absolute value of the time interval between the corresponding farming activities, and obtains the time difference value. At the same time, the system calculates the absolute difference value of the farming uniformity index of adjacent plots, and obtains the uniformity difference value. In order to eliminate the dimension influence of different difference types, the system normalizes the time difference value and the uniformity difference value respectively, divides the time difference value by the maximum change range of farming activity time in the research area, and divides the uniformity difference value by the maximum change range of farming uniformity, so that both types of difference values are converted to the 0-1 interval. Then the system determines the time difference weight and the uniformity difference weight according to the agricultural expert experience and statistical analysis results, usually sets the time difference weight to 0.6-0.7, and sets the uniformity difference weight to 0.3-0.4. Finally, the system multiplies the normalized time difference value by the time difference weight, multiplies the normalized uniformity difference value by the uniformity difference weight, and then adds the two products to obtain the farming activity difference index.

[0063] In some embodiments, the calculation and optimization of the farming activity difference index can be achieved in various ways: optionally, a multi-level adjacency relationship analysis method is used to improve the spatial representativeness of the synchronization evaluation, the system not only considers the directly adjacent plots, but also analyzes the second and third adjacent plots with similar distances, calculates the weighted average difference index of different adjacent levels, obtains a more comprehensive and stable synchronization evaluation result, adjusts the adjacency weight according to the size and shape characteristics of the plot, ensures that the adjacency relationship analysis of large plots and irregular plots is more reasonable, and finally generates a comprehensive farming activity difference index with spatial hierarchy; optionally, a dynamic weight optimization mechanism based on machine learning is established, the system collects historical field verification data of contract right changes, trains the weight optimization model, automatically learns the optimal configuration scheme of the time difference weight and the uniformity difference weight, uses genetic algorithm or particle swarm algorithm for weight parameter optimization, considers the weight difference under different crop types, different geographical environments and different management modes, establishes an adaptive weight adjustment strategy under multiple scenarios, and improves the accuracy and applicability of the farming activity difference index. It can be understood that other ways can also be used to realize the quantitative evaluation and index calculation of farming activity synchronization, which is not limited here.

[0064] In some embodiments, the step specifically includes the following steps: The time span of the farming activity time of each plot in the study area is calculated, the time difference value is divided by the time span to obtain the normalized time difference value, the change range of the tillage uniformity of each plot in the study area is calculated, the uniformity difference value is divided by the change range to obtain the normalized uniformity difference value, the time difference weight and the uniformity difference weight are determined based on the indication degree of farming activity on land contract right change, the normalized time difference value and the time difference weight, and the normalized uniformity difference value and the uniformity difference weight are multiplied and added respectively to obtain the farming activity difference index.

[0065] The farming activity time of each plot in the study area refers to the set of time points of key agricultural operations of all farmland plots in the study area. The time span is the difference between the maximum and minimum values of the farming activity time in the region, reflecting the overall range of the farming activity time distribution. The time difference value is the absolute difference value of the farming activity time between adjacent plots. The normalized time difference value is a standardized time difference index that eliminates the influence of time dimension. The change range of tillage uniformity is the difference between the maximum and minimum values of the tillage uniformity of each plot in the region. The uniformity difference value is the absolute difference value of the tillage uniformity between adjacent plots. The normalized uniformity difference value is a standardized uniformity difference index. The indication degree of farming activity on land contract right change reflects the contribution strength of different farming characteristics to the identification of contract right change. The time difference weight and the uniformity difference weight are weight coefficients that quantify the importance of the two types of characteristics. The farming activity difference index is a quantitative index for comprehensive evaluation of the coordination degree of farming activities of adjacent plots.

[0066] When performing this step, first, collect the time data of farming activities for all plots in the study area, and establish a time dataset T_region = {T1, T2,..., Tn}, where Ti represents the time vector of farming activities for the ith plot, and n is the total number of plots. Calculate the regional distribution range of each type of farming activity time: for sowing time, calculate Span_seed = max(T_seed) - min(T_seed); for fertilizing time, calculate Span_fertilize = max(T_fertilize) - min(T_fertilize); for harvesting time, calculate Span_harvest = max(T_harvest) - min(T_harvest). Establish a time span matrix S = [Span_seed, Span_fertilize, Span_harvest,...], covering all key farming activity types. For adjacent plots i and j, calculate the time difference value of each farming activity: ΔT_activity = |Ti_activity - Tj_activity|, and then perform normalization processing: Normalized_ΔT_activity = ΔT_activity / Span_activity, to ensure that the normalized value range is within the interval [0, 1]. Collect the uniformity data of each plot in the study area, and establish a uniformity dataset U_region = {U1, U2,..., Un}, calculate the range of uniformity: Range_uniformity = max(U_region) - min(U_region). For the uniformity difference value ΔU = |Ui - Uj| of adjacent plots, perform normalization processing: Normalized_ΔU = ΔU / Range_uniformity. Based on historical contract right change case analysis and agricultural expert knowledge, determine the weight coefficient: the indication strength of time difference on contract right change is quantified by calculating the discrimination degree of historical changed plots and unchanged plots in time difference, and the information gain method is used to calculate the time difference weight W_time, which is usually set to 0.6-0.7; the indication strength of uniformity difference is calculated by a similar method, and the uniformity difference weight W_uniformity is usually set to 0.3-0.4, ensuring that W_time + W_uniformity = 1. The specific formula for calculating the farming activity difference index is: Synchronization_Index = W_time × Σ(Normalized_ΔT_activity) / N_activities + W_uniformity × Normalized_ΔU, where N_activities is the total number of farming activity types.To improve calculation accuracy, a weighted average method is used to handle the time differences of various agricultural activities: Weighted_Time_Diff = Σ(αk × Normalized_ΔT_k), where αk is the importance weight of the k-th agricultural activity, with sowing and harvesting typically weighted at 0.4, and fertilization and irrigation weighted at 0.2. The final synchronization index calculation formula is: Synchronization_Index = W_time × Weighted_Time_Diff + W_uniformity × Normalized_ΔU. A quality control mechanism for the synchronization index is established, and numerical correction is performed when the calculation result exceeds the range of [0,1]. This generates a synchronization index matrix for each plot pair, providing a standardized quantitative basis for subsequent identification of contract rights changes.

[0067] S105. When the agricultural activity difference index of a target plot is greater than the preset difference threshold, the target plot is marked as a suspected plot for change of contract rights.

[0068] In this context, "target plot" refers to a specific farmland unit within the study area that requires identification of changes in contractual rights. These plots are typically defined based on plot boundaries established through cadastral management or agricultural censuses. An agricultural activity difference index exceeding a preset difference threshold indicates a significant difference in agricultural activity patterns between the plot and its neighboring plots. This difference may be caused by factors such as changes in management entities, alterations in farming methods, or transfers of contractual rights. The preset difference threshold is a critical numerical standard used to determine whether a plot is suspected of having undergone a change in contractual rights. This threshold is usually determined by analyzing historical cases of contractual rights changes and the range of variation in normal agricultural activities. "Suspected contractual rights change plots" represents a set of plots initially selected as potentially susceptible to contractual rights changes. These plots require further verification to confirm whether a genuine change in contractual rights has occurred. The labeling process refers to the technical process by which the system automatically identifies and records plots that meet the criteria, establishing a candidate target library for subsequent in-depth analysis.

[0069] Specifically, the system first calculates the difference index of agricultural activities for each plot in the study area and compares it with the threshold value. The system establishes a dynamic threshold determination mechanism, sets a reasonable synchronization threshold range based on the agricultural production characteristics, crop planting structure and historical contract right change mode of the study area, and usually sets the threshold value between 0.6 and 0.8. When the difference index of agricultural activities of a plot exceeds the preset threshold value, the system will automatically add the plot to the list of suspected changed plots and record its geographic location, plot number, difference index value and specific agricultural activity type that exceeds the threshold value. The system also generates a spatial distribution map of suspected changed plots, marking the difference index level and risk level of each suspected plot, providing intuitive visual monitoring results for management departments. To avoid misjudgment, the system sets multiple verification mechanisms, including time stability test of difference index, correlation analysis of adjacent plots and seasonal change mode check, to ensure the high credibility of the marked suspected plots. At the same time, the system establishes a dynamic update mechanism for suspected plots to continuously monitor and adjust the list of suspected plots based on newly acquired remote sensing data.

[0070] In some embodiments, the identification and marking of suspected contract right change plots can be achieved in various ways: optionally, a multi-threshold grading marking strategy is adopted to improve the refinement of identification, the system sets high, medium and low synchronization threshold levels corresponding to high-risk, medium-risk and low-risk suspected change levels, respectively, and optimizes the allocation of monitoring resources through a grading management mechanism. High-risk plots are given priority for on-site verification, medium-risk plots are subjected to enhanced remote sensing monitoring frequency, and low-risk plots are included in the regular monitoring range. At the same time, a dynamic risk level adjustment mechanism is established to continuously optimize the grading standards based on subsequent monitoring results and verification feedback; optionally, a suspected plot identification method based on spatio-temporal clustering is established, the system not only considers the difference index of a single plot, but also analyzes the spatial clustering pattern and temporal occurrence regularity of suspected plots, uses spatio-temporal clustering algorithm to identify hotspots of contract right change and concentrated time periods, and improves the accuracy of individual plot identification through regional change pattern analysis. At the same time, a neighborhood verification mechanism is established, when a plot is marked as suspected change, the synchronization change of its surrounding plots is automatically checked, and spatial correlation analysis is used to reduce isolated misjudgment. It can be understood that other ways can also be used to achieve intelligent identification and classification marking of suspected contract right change plots, which are not limited here.

[0071] S106、When the correlation coefficient between the time series curve of the difference index of agricultural activities of the suspected contract right change plot and the historical time series curve is less than the preset correlation threshold value, the suspected contract right change plot is determined as a contract right change plot.

[0072] The time series curve of the farming activity difference index represents the trajectory of the farming activity difference index of the suspected land plot over time in the current monitoring period, and can reflect the time evolution characteristics of the farming activity pattern of the land plot. The historical time series curve is the time variation pattern of the farming activity difference index of the land plot in historical normal years, which serves as a benchmark for determining whether there is an abnormal change in the current year. The correlation coefficient is used to represent the linear correlation between the current time series curve and the historical time series curve, and the numerical range is between -1 and 1. The closer the correlation coefficient is to 1, the more similar the two curves are, and the closer it is to 0, the weaker the correlation is. The preset correlation threshold is a critical standard for determining whether the similarity of the time series curves reaches a normal level, and is usually set between 0.7 and 0.8. The contract right change land plot is used to represent the farmland plot that has truly undergone contract right change after multiple verification, and these plots need to be included in the scope of key supervision and right ownership verification. The determination process is the final identification process for screening out the true change land plot from the suspected land plot through time series pattern comparison analysis.

[0073] Specifically, the system first extracts the historical farming activity difference index data of the suspected contract right change land plot, and usually selects the data of stable years in the past 3-5 years as the historical benchmark. The system will statistically analyze the time series data of the difference index of multiple years, calculate the mean and standard deviation of each time node, and construct a standard historical time series curve template. Then the system extracts the time series data of the farming activity difference index of the land plot in the current year, and generates the time series curve of the current year. In the correlation coefficient calculation process, the system uses the Pearson correlation coefficient method to compare the numerical values of the corresponding time nodes of the current time series curve and the historical time series curve, and calculates the linear correlation degree of the two curves. The specific calculation formula is: r = Σ[(Xi-X̄)(Yi-Ȳ)] / √[Σ(Xi-X̄)²Σ(Yi-Ȳ)²], where Xi represents the value of the historical time series curve at the i-th time point, Yi represents the value of the current time series curve at the i-th time point, X̄ and Ȳ represent the mean values of the two curves, respectively. When the calculated correlation coefficient is less than the preset correlation threshold, the system determines that the farming activity pattern of the land plot has changed significantly, and promotes it from the suspected land plot to the confirmed contract right change land plot. The system also generates a detailed change confirmation report, including the correlation coefficient value, the time series curve comparison chart and the change confidence evaluation.

[0074] In some embodiments, the final confirmation and verification of the contract right changed plot can be achieved in various ways: optionally, a multi-dimensional time series similarity analysis method is used to improve the accuracy of change identification. The system not only calculates the time correlation of the agricultural activity difference index, but also analyzes the correlation characteristics of multiple dimensions such as vegetation index time series, tillage uniformity time series, and agricultural activity time series. Through comprehensive comparison of multi-dimensional time series patterns, a more comprehensive and reliable change determination mechanism is established. At the same time, the dynamic time warping algorithm is used to process the time shift problem of the time series curve, improving the accuracy of time series pattern comparison between different years. Finally, the multi-index fusion decision determines the confirmation result of the contract right change. Optionally, a change pattern recognition system based on machine learning is established. The system collects a large number of historical contract right change verification cases and corresponding time series data, trains the contract right change identification model, and uses support vector machine, random forest or deep neural network algorithm to learn the time series feature pattern of the changed plot, automatically identifies and classifies different types of contract right change. At the same time, a model performance evaluation and continuous optimization mechanism is established to continuously improve the identification accuracy and generalization ability of the model based on new verification data, realizing intelligent identification and classification of contract right change. It can be understood that other ways can also be used to achieve accurate identification and reliable verification of contract right changed plots, which are not limited here.

[0075] In some embodiments, the step specifically includes: The time series data of the agricultural activity difference index of the suspected contract right changed plot in multiple historical years without contract right change is obtained, the time series data of the agricultural activity difference index in multiple historical years is averaged to obtain a baseline historical time series curve, the agricultural activity difference index of each agricultural activity period in the current year of the suspected contract right changed plot is collected to generate a current time series curve, the correlation coefficient of the current time series curve and the baseline historical time series curve is calculated, and when the correlation coefficient is less than a preset correlation threshold, the suspected contract right changed plot is determined as a contract right changed plot.

[0076] The suspected contract right changed plot refers to a plot unit with an abnormal agricultural activity difference index identified through preliminary screening. The historical year without contract right change refers to a past period when the contract right ownership of the plot is stable and the management subject has not changed. The time series data of the agricultural activity difference index is a numerical sequence arranged by time for the difference index of each agricultural activity period in a specific year. The baseline historical time series curve is the statistical average trajectory of the difference index time series data in multiple historical years, representing the normal agricultural activity pattern of the plot. The current time series curve is the change trajectory of the agricultural activity difference index of the suspected plot in the current monitoring year. The correlation coefficient quantifies the linear correlation between the current time series curve and the baseline historical time series curve. The preset correlation threshold is the critical standard for determining the similarity of time series patterns. The contract right changed plot is a farmland unit that has been verified and confirmed to have changed contract right.

[0077] When performing this step, first extract the difference index time series data of the suspected contract right change plot for the past 3-5 stable years from the historical database. Establish a historical data matrix H = [H1, H2,..., Hm], where Hi represents the time series data vector of the i-th historical year, and m is the total number of historical years. The time series data of each year contains the difference index of the key farming activity period: Hi = [Si1, Si2,..., Sip], where Sij represents the difference index of the j-th farming activity period in the i-th year, and p is the total number of farming activity periods. Perform data quality checks and remove years with more than 20% missing data, and use linear interpolation to complete the missing data points. Time alignment is performed on the historical multi-year data to ensure that the data of the same farming activity period corresponds to the same time node. Calculate the baseline historical time series curve: Baseline(j) = Σ(Sij) / m, where j is the farming activity period number, and obtain the baseline time series vector B = [Baseline(1), Baseline(2),..., Baseline(p)]. Calculate the standard deviation σj = √[Σ(Sij-Baseline(j))² / (m-1)] at each time node, and establish the confidence interval [Baseline(j)-2σj, Baseline(j)+2σj] as the normal change range. Extract the difference index of each farming activity period in the current year for the suspected plot, and construct the current time series curve C = [C1, C2,..., Cp], where Cj is the difference index of the j-th farming activity period in the current year. Calculate the time series correlation using the Pearson correlation coefficient formula: r = Σ[(Cj-C̄)(Baseline(j)-B̄)] / √[Σ(Cj-C̄)²Σ(Baseline(j)-B̄)²], where C̄ and B̄ are the mean values of the current time series and the baseline time series, respectively. Introduce lag correlation analysis to handle time offset problems, calculate the correlation coefficient rk = Corr(C(t), B(t+k)) at different lag periods k, and select the maximum correlation coefficient as the final result. Set a pre-set correlation threshold r_threshold, which is usually determined based on ROC analysis of historical validation data, with a value range of 0.7-0.8. When the calculated correlation coefficient r < r_threshold, it is determined that the farming activity pattern of the plot has deviated significantly, and it is confirmed as a contract right change plot from the suspected state. Establish a confidence evaluation mechanism for change confirmation: Confidence = (r_threshold-r) / r_threshold, the higher the confidence, the more reliable the change determination. Generate a detailed change confirmation report, including plot number, correlation coefficient value, confidence level, time series comparison chart and abnormal period identification, record the change confirmation time and analysis parameters, and establish a contract right change plot archive to provide scientific basis for subsequent ownership verification and management decision-making.

[0078] The method provided by the embodiment is further described in detail below. Please refer to Figure 2 , another flowchart of the method for identifying land contract right change based on remote sensing images in the embodiment of the present application.

[0079] S201, when the correlation coefficient of the time series curve of the agricultural activity difference index of the suspected contract right change plot and the historical time series curve is less than a preset threshold, the suspected contract right change plot is determined as a contract right change plot.

[0080] The suspected contract right change plot refers to a plot with an abnormal agricultural activity difference index identified through preliminary screening. The time series curve of the agricultural activity difference index is a data sequence composed of the difference index of the plot at each time node in the current monitoring period. The historical time series curve is a time variation benchmark mode of the difference index of the plot in the historical stable years. The correlation coefficient quantifies the linear correlation degree of the two time series curves, and the numerical range is -1 to 1. The preset threshold is a critical standard for determining the similarity of the time series mode. The contract right change plot is a farmland unit that has been verified to have a contract right change.

[0081] When performing this step, first, the agricultural activity difference index data of the suspected plot in the past 3-5 stable years is extracted, and the historical benchmark data set is formed by arranging the data in chronological order. The historical multi-year data is time-aligned to ensure that the data of the same agricultural activity period correspond to the same time node. The mean value of the corresponding time node of each year is calculated to construct the standard historical time series curve H(t), where t represents the agricultural activity period number. The time series data of the agricultural activity difference index of the suspected plot in the current year is extracted to form the current time series curve C(t). The correlation between the two curves is calculated using the Pearson correlation coefficient formula: r = Σ[(C(t)-C̄)(H(t)-H̄)] / √[Σ(C(t)-C̄)²Σ(H(t)-H̄)²], where C̄ and H̄ are the mean values of the current curve and the historical curve, respectively. When the correlation coefficient r is less than the preset threshold (usually 0.75), it is determined that the agricultural activity mode of the plot has changed significantly, and the plot is promoted from the suspected state to the confirmed contract right change plot. A change confirmation record is generated, including the plot number, the correlation coefficient value, the change confidence, and the time series comparison chart, and a contract right change plot database is established for subsequent management.

[0082] S202, obtaining the time series curve of the agricultural activity difference index of the plot around the contract right change plot.

[0083] The contract right change plot refers to the farmland unit where the contract right change has been confirmed. The surrounding plots are the farmland plots that are adjacent to or close to the contract right change plot in spatial position. The difference index time series curve of agricultural activities is the data trajectory of the difference index of each surrounding plot changing with time during the monitoring period. The acquisition process involves the technical process of spatial neighborhood analysis and time series data extraction.

[0084] When performing this step, first, a spatial buffer centered on the contract right change plot is established, and the buffer radius is usually set to 500-1000 meters, ensuring that the neighborhood range affected by the change is covered. All farmland plots within the buffer are identified using spatial analysis algorithms, and a spatial adjacency relationship matrix is established. The identified surrounding plots are numbered and attribute records are recorded, including plot area, boundary coordinates, distance and azimuth angle from the center plot, and other information. The historical data of the difference index of agricultural activities of each surrounding plot are extracted from the remote sensing data processing system, and the time span is consistent with the analysis period of the contract right change plot. The original data extracted are subjected to quality inspection, and abnormal values, missing values and other data quality problems are identified and processed. Time interpolation method is used to complete the missing time series node data, ensuring that the time series curves of all surrounding plots have the same time resolution and data integrity. The time series data of the difference index of agricultural activities of each surrounding plot are standardized to eliminate the baseline differences between different plots. A time series curve database of surrounding plots is established, and each plot is assigned a unique identifier, recording its spatial relationship with the contract right change plot and the time series data characteristics. The distribution map of surrounding plots and the set of time series curves are generated, providing a data basis for subsequent similarity analysis and regional division.

[0085] S203, according to the similarity of the difference index time series curve of agricultural activities of the contract right change plot and the surrounding plots, the surrounding plots are divided into synchronous change region and non-synchronous change region.

[0086] The contract right change plot and the surrounding plots are a set of spatially related farmland units. The difference index time series curve of agricultural activities quantifies the consistency of agricultural activity patterns between different plots. The synchronous change region refers to the plot set that presents similar changes in agricultural activity patterns with the contract right change plot. The non-synchronous change region refers to the plot region where the agricultural activity pattern is significantly different from the contract right change plot. The division process realizes regional classification through similarity calculation and threshold determination.

[0087] When performing this step, first extract the time series curve of the farming activity difference index of the contracted management right changed plot as a reference template T(t). Calculate the similarity index of each surrounding plot time series curve N(t) and the reference template one by one. Comprehensive evaluation is carried out by using multiple similarity measurement methods: calculate the Pearson correlation coefficient to evaluate the linear correlation, calculate the Euclidean distance to evaluate the numerical difference degree, and use the dynamic time warping algorithm to process the time series shift problem. The similarity comprehensive score S is calculated by weighted fusion: S = w1xr + w2x(1-d / dmax) + w3xDTW, where r is the correlation coefficient, d is the standardized Euclidean distance, DTW is the dynamic time warping similarity, w1, w2, w3 are weight coefficients and satisfy w1+w2+w3=1. Set the similarity threshold St (usually 0.8), when the comprehensive similarity S of the surrounding plot is ≥ St, the plot is divided into the synchronous change area; when S< St, it is divided into the non-synchronous change area. Spatial connectivity analysis is carried out on the division result, adjacent areas of the same type are merged, and spatial fragmentation phenomenon is eliminated. Generate regional division thematic map, identify synchronous change area and non-synchronous change area with different colors. Calculate the area statistics and distribution characteristics of each region, analyze the spatial influence range of contracted management right change. Establish the regional division result database, record the regional attribution, similarity value and spatial position information of each plot, and provide basic data support for the centralized and continuous analysis of contracted management right change.

[0088] S204, calculate the spatial autocorrelation coefficient of the farming activity difference index of the plot in the synchronous change area, and determine the centralized and continuous area of the contracted management right change according to the spatial autocorrelation coefficient.

[0089] The synchronous change area refers to the plot set that presents similar farming activity change mode with the contracted management right changed plot. The farming activity difference index is a numerical index for quantifying the coordination degree of farming activities between plots. The spatial autocorrelation quantifies the correlation degree of attribute values of adjacent plots in geographical space, and reflects the aggregation characteristics of spatial phenomena. The centralized and continuous area refers to the combination of plots that are continuously distributed in space and have similar change characteristics.

[0090] When performing this step, first, the spatial weight matrix W of each plot in the synchronous change area is established, and the adjacency criterion is used to define the spatial relationship, and when two plots share a boundary, the weight value is set to 1, otherwise 0. The difference index value of each plot in the region is extracted, and the attribute vector X = [x1, x2, …, xn] is formed, where xi represents the difference index value of the ith plot. The global Moran's I spatial autocorrelation coefficient is calculated: I = (n / S0) x [∑i∑j wij(xi-x̄)(xj-x̄)] / [∑i(xi-x̄)²], where n is the total number of plots, S0 is the sum of all elements of the weight matrix, wij is the spatial weight between plots i and j, x̄ is the mean of the difference index. The local Moran's Ii coefficient is calculated to identify the spatial hot spot area: Ii = (xi-x̄) x ∑j wij(xj-x̄) / σ², where σ² is the variance of the difference index. Significance test is performed, and the Z statistic is calculated: Z = (I-E[I]) / √Var[I], where E[I] is the expected value, and Var[I] is the variance. When the Z value is greater than 1.96 and the p value is less than 0.05, it is determined that there is a significant positive spatial autocorrelation. Identify the combination of plots with a local Moran's I significantly positive and a difference index higher than the mean, and determine it as a concentrated contiguous area of high-high aggregation type. Using the connectivity analysis algorithm, the high aggregation plots that are spatially adjacent are merged to form the final concentrated contiguous area of contract right change. The area, perimeter, compactness and other geometric characteristics of each contiguous area are calculated to generate the distribution map and statistical report of the contiguous area.

[0091] S205, establish the time series feature vector of the difference index of the contract right change plot, which includes fluctuation amplitude, periodicity and trend.

[0092] The contract right change plot is a farmland unit that has been confirmed to have a contract right change. The time series of agricultural activity difference index is a data sequence of the change of the difference index of the plot over time during the monitoring period. The time series feature vector is a combination of multi-dimensional numerical values that describe the statistical characteristics and change rules of the time series data. The fluctuation amplitude quantifies the variation degree of the time series data. The periodicity reflects the repetitive change pattern of the time series data. The trend describes the long-term change direction of the time series data.

[0093] When performing this step, first extract the agricultural activity difference index time series data Y(t) of the contracted right changed plot, where t is the time node number. Calculate the fluctuation amplitude characteristics: use the standard deviation method σ = √[Σ(Y(t)-Ȳ)² / (n-1)] to quantify the data dispersion degree, use the coefficient of variation CV = σ / Ȳ to eliminate the influence of mean value, calculate the range R = max(Y(t))-min(Y(t)) to reflect the change range, use the mean absolute deviation MAD = Σ|Y(t)-Ȳ| / n to evaluate the fluctuation intensity. Calculate the periodicity characteristics: use the fast Fourier transform FFT to analyze the frequency domain characteristics of the time series, identify the main periodic components, calculate the power spectral density PSD to determine the contribution degree of each frequency component, use the autocorrelation function ACF(k) = Σ[Y(t)×Y(t+k)] / n to detect the periodic pattern with lag k period, determine the frequency and amplitude of the significant period through spectral analysis. Calculate the trend characteristics: use the linear regression method to fit the trend line Y(t) = at + b, where the slope a represents the trend direction and intensity, calculate the determination coefficient R² to evaluate the trend fitting degree, use the Mann-Kendall non-parametric test to evaluate the trend significance, calculate the Theil-Sen slope estimator to obtain the robust trend estimate. Construct the time series feature vector F = [σ, CV, R, MAD, main period frequency, main period amplitude, ACF maximum value, trend slope, trend R², MK statistic], form the time series feature benchmark template of the contracted right changed plot. Standardize the feature vector to eliminate the dimensional difference between different features, provide standardized feature representation for subsequent similarity calculation.

[0094] S206, calculate the similarity coefficient between the agricultural activity difference index time series feature vector of the surrounding plot and the time series feature vector of the contracted right changed plot.

[0095] The surrounding plot is the farmland unit within the neighborhood range of the contracted right changed plot. The agricultural activity difference index time series feature vector is a multidimensional numerical combination describing the time series change characteristics of the plot. The similarity coefficient quantifies the degree of similarity between the two feature vectors.

[0096] When performing this step, first extract the time series data of the difference index of agricultural activities of each surrounding plot, and calculate the time series feature vector according to the same method of the contract right changed plot. Let the feature vector of the contract right changed plot be F_ref=[f1_ref,f2_ref,...,fm_ref], and the feature vector of the i-th surrounding plot be F_i=[f1_i,f2_i,...,fm_i], where m is the feature dimension. A variety of similarity measurement methods are used: calculate the cosine similarity cos_sim=(F_ref·F_i) / (||F_ref||×||F_i||) to evaluate the similarity of the vector angle; calculate the Euclidean distance similarity eucl_sim=1 / (1+√Σ(fj_ref-fj_i)²) to evaluate the distance similarity in the feature space; calculate the Pearson correlation coefficient corr_coef=Σ[(fj_ref-f̄_ref)(fj_i-f̄_i)] / √[Σ(fj_ref-f̄_ref)²Σ(fj_i-f̄_i)²] to evaluate the linear correlation. Since different features have different indicating effects on the contract right change, a weighted similarity calculation method is designed: weighted_sim=Σwj×sim_j(fj_ref,fj_i), where wj is the weight of the j-th feature, and sim_j is the similarity function of the corresponding feature. The weight allocation is based on the feature importance: the fluctuation amplitude feature weight is 0.3, the periodicity feature weight is 0.4, and the trend feature weight is 0.3. Calculate the comprehensive similarity coefficient: S_total=α×cos_sim+β×eucl_sim+γ×corr_coef+δ×weighted_sim, where α, β, γ, δ are fusion weights and satisfy α+β+γ+δ=1. The optimal weight configuration is determined by cross-validation, and usually α=0.25, β=0.25, γ=0.25, δ=0.25. Generate a similarity coefficient matrix to record the similarity values of each surrounding plot and the contract right changed plot, and establish a similarity ranking list to provide quantitative basis for subsequent regional division and pattern recognition.

[0097] S207, determine the time series feature similarity score according to the similarity coefficient, which reflects the consistency degree of the change of agricultural activities.

[0098] The similarity coefficient is a numerical indicator that quantifies the similarity degree of two time series feature vectors. The time series feature similarity score is a standardized evaluation index calculated based on the similarity coefficient. The consistency degree of the change of agricultural activities refers to the level of coordination and unity of different plots in terms of the time, intensity and mode of agricultural activities.

[0099] When performing this step, first, the calculated multiple similarity coefficients are standardized and fused. The cosine similarity, Euclidean distance similarity, Pearson correlation coefficient, and weighted similarity are unified in numerical range to ensure that all similarity coefficients are mapped to the [0, 1] interval. The time series feature similarity score is calculated using a comprehensive score function: Score = w1 x cos_sim_norm + w2 x eucl_sim_norm + w3 x corr_coef_norm + w4 x weighted_sim_norm, where w1, w2, w3, and w4 are weight coefficients of each similarity index, and norm represents the normalized numerical value. The weight distribution is based on the contribution of each index to the detection of farming activity changes: cosine similarity weight 0.2 (reflecting overall pattern similarity), Euclidean distance similarity weight 0.25 (reflecting numerical difference), Pearson correlation coefficient weight 0.3 (reflecting linear relationship), and weighted similarity weight 0.25 (reflecting feature importance). A confidence adjustment mechanism is introduced to correct the similarity score based on the integrity and quality of the time series data: Adjusted_Score = Score x Quality_Factor, where Quality_Factor is calculated based on data missing rate, abnormal value proportion, and time series length. A statistical distribution model of similarity score is established to calculate the mean, standard deviation, and quantile of the score, providing statistical basis for threshold setting. The similarity score distribution histogram and cumulative distribution function are generated to analyze the distribution characteristics of the score in different numerical intervals. The final similarity scores of the surrounding plots are recorded to establish a database of plot numbers and scores, providing a quantitative basis for subsequent regional division decisions.

[0100] S208, the surrounding plots with time series feature similarity scores greater than the set threshold are classified into the synchronous change region, and the surrounding plots with time series feature similarity scores less than or equal to the set threshold are classified into the non-synchronous change region.

[0101] The time series feature similarity score is a quantitative index reflecting the consistency of farming activity changes between plots. The threshold is set as the critical standard for distinguishing between synchronous and non-synchronous change plots. The synchronous change region refers to a collection of plots that exhibit similar farming activity change patterns with the plot of contract right change. The non-synchronous change region refers to a plot area where the farming activity pattern is significantly different from the plot of contract right change.

[0102] In this step, the similarity scoring threshold is first determined based on historical contractual right change verification data and expert experience. Using ROC curve analysis, plots with known historical contractual right changes are considered positive samples, while plots without changes are considered negative samples. The true positive rate and false positive rate are calculated under different thresholds, and the threshold maximizing the AUC value is selected as the optimal classification standard, typically set between 0.75 and 0.85. A dynamic threshold adjustment mechanism is established to optimize the threshold based on the differences in agricultural production characteristics, crop types, and management models in the study area. The similarity scores of each surrounding plot are compared with the set threshold: when Score_i > Threshold, plot i is classified into the synchronous change area, and its plot number, coordinates, similarity score, and spatial distance from the contractual right change plot are recorded; when Score_i ≤ Threshold, plot i is classified into the asynchronous change area, and its degree of difference and possible influencing factors are noted. Spatial connectivity analysis is performed on the classification results to identify connected components within the synchronous change area, merge adjacent similar plots, and eliminate spatial fragmentation. Calculate the geometric characteristic indicators of each region, including parameters such as area, perimeter, shape index, and compactness. Generate thematic maps of regional division, using different colors and symbols to identify synchronously changed and asynchronously changed regions, and annotate the similarity scores of each plot. Establish a data table of regional division results, recording the regional affiliation, score ranking, spatial location, and adjacency relationships of each plot, and statistically analyzing the total area, number of plots, and spatial distribution characteristics of synchronously changed regions, providing basic data for assessing the impact scope of contractual rights changes.

[0103] S209. Obtain the on-site verification results of land parcels with changes in contracted rights during historical periods, and establish a correspondence between the time series curve of the agricultural activity difference index and the on-site verification results.

[0104] Historical period refers to the time period during which monitoring and verification of changes in contracted rights have been completed. The on-site verification results of plots with changed contracted rights are records of the actual changes in contracted rights obtained through on-site investigations. The time-series curve of the agricultural activity difference index is the data trajectory of the difference index of the plot over time during the monitoring period. Correspondence refers to the correlation mapping pattern between the characteristics of the time-series curve and the on-site verification results.

[0105] When performing this step, first collect field verification data in the study area for 3-5 years, including the accurate location of the plot of land where the contract right is changed, the time of change, the type of change, the reason for change, and the information of the new contract subject. Establish a field verification database to record the basic attributes of each verification plot: plot number, geographic coordinates, area size, original contract holder information, new contract holder information, change time, change type (subcontracting, leasing, exchange, inheritance, etc.), and verification confirmation time. Extract the time series data of the difference index of agricultural activities of each verification plot in the corresponding historical period, ensuring that the time span of the time series data covers at least one complete agricultural production cycle before and after the change of contract right. Perform quality inspection and preprocessing on the time series data, eliminate outliers and missing data, and use interpolation methods to complete the data gap, ensuring the continuity and integrity of the time series data. Establish an association data structure between time series curves and verification results, match the time series curve characteristics of each verification plot with its change type and change time. Analyze the time series curve feature patterns corresponding to different change types: subcontracting changes usually show sudden changes in agricultural activity time and step changes in difference index; leasing changes show a gradual downward trend in difference index; exchange changes show a sharp fluctuation in difference index in the short term and then tend to be stable. Use statistical analysis methods to quantify the association strength between time series characteristics and change results, calculate the prediction accuracy and recall rate of different time series patterns for each type of contract right change. Establish a time series feature-change type correspondence database to record the characteristic parameters of typical time series patterns, the corresponding change types, and the confidence level, providing reference templates and decision-making basis for determining the change type of newly identified plots.

[0106] S210, according to the corresponding relationship, identify the agricultural activity feature patterns of different types of land contract right changes, and determine the change type of the newly identified contract right change plot based on the agricultural activity feature patterns.

[0107] The corresponding relationship refers to the association mapping pattern between the time series curve characteristics of the difference index of agricultural activities and the field verification results of the contract right changes. Different types of land contract right changes include subcontracting, leasing, exchange, inheritance, and termination of circulation, among others. Agricultural activity feature patterns are the typical change rules and numerical feature combinations of each type of contract right change on the time series curve. Change type refers to the specific form and nature classification of the contract right change.

[0108] When performing this step, first, based on historical field verification data and corresponding time series curve data, cluster analysis and pattern recognition methods are used to extract typical patterns of each type of contract right change. For the transfer type of change, the analysis found that the characteristic pattern is that the difference index of agricultural activities decreases significantly at the change time point. The difference index before the change usually maintains a stable level above 0.8, and after the change, it decreases to the interval of 0.4-0.6 within 1-2 agricultural activity cycles, and then gradually stabilizes at a new value level. Calculate the quantitative characteristic parameters of the transfer mode: the mean difference of the difference index before and after the change Δμ = μ_before-μ_after is usually greater than 0.3, the standard deviation σ_change within the change time window is usually greater than 0.15, and the absolute value of the time series curve slope |slope| near the change point is usually greater than 0.08 / cycle. For the rental type of change, the characteristic pattern is a gradual downward trend in the difference index, with a change process lasting 2-3 agricultural activity cycles and eventually stabilizing at a lower value level. Calculate the characteristic parameters of the rental mode: the linear trend slope is usually in the interval of -0.03 to -0.06 / cycle, the correlation coefficient R² of the decline process is usually greater than 0.7, and the stable value of the difference index after the change is usually less than 0.5. For the exchange type of change, the characteristic pattern is a sharp fluctuation in the difference index in the short term, followed by a rapid recovery to near the original level. Calculate the characteristic parameters of the exchange mode: the coefficient of variation CV during the fluctuation is usually greater than 0.25, the fluctuation range range = max-min is usually greater than 0.4, the fluctuation duration is usually 1-2 agricultural activity cycles, and the difference between the difference index after recovery and before the change is usually less than 0.1. Establish a decision tree model for change type recognition: first, determine whether there is a significant trend change in the time series curve. If the linear trend slope |slope| < 0.02 and R² < 0.5, it is determined as a stable mode without contract right change; if slope < -0.025 and R² > 0.6, it is determined as a rental type of change; if there is a significant step change point and the mean difference Δμ before and after the change is greater than 0.25, it is determined as a transfer type of change; if CV > 0.2 in the short term and the difference between the final stable value and the initial value is less than 0.15, it is determined as an exchange type of change. For the newly identified contract right change plot, extract its agricultural activity difference index time series curve, calculate the corresponding characteristic parameters, and determine the change type through the decision tree model. Establish a confidence evaluation mechanism to calculate the recognition confidence according to the matching degree of the new plot time series characteristics and the typical pattern: confidence = 1 - |feature_new-feature_typical| / feature_range, where feature_range is the range of the feature in the historical sample.The change type identification report is generated, including plot number, identified change type, confidence level, key feature parameter value and time sequence curve comparison chart, to provide technical basis and priority ranking for on-site verification and ownership confirmation of the management department.

[0109] The land contract right change identification system in the embodiment of the application is described from the perspective of hardware processing below. Please refer to Figure 3 FIG. 1 is a schematic diagram of an entity device structure of the land contract right change identification system in the embodiment of the application.

[0110] It should be noted that, Figure 3 The structure of the land contract right change identification system shown is only an example, and should not bring any limitation to the function and use range of the embodiment of the application.

[0111] As Figure 3 shown, the land contract right change identification system includes a central processing unit (CPU) 301, which can perform various appropriate actions and processes according to programs stored in a read-only memory (ROM) 302 or programs loaded from a storage portion 308 to a random access memory (RAM) 303, such as performing the method described in the above embodiment. In the RAM 303, various programs and data required for system operation are also stored. The CPU 301, the ROM 302 and the RAM 303 are connected to each other through a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0112] The following components are connected to the I / O interface 305: an input portion 306 including an audio input device, a button switch, etc.; an output portion 307 including a liquid crystal display (LCD) and an audio output device, an indicator light, etc.; a storage portion 308 including a hard disk, etc.; and a communication portion 309 including a network interface card such as a LAN (Local Area Network) card, a modem, etc. The communication portion 309 performs communication processing via a network such as the Internet. A drive 310 is also connected to the I / O interface 305 as needed. A removable medium 311 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc. is mounted on the drive 310 as needed, so that a computer program read therefrom is installed in the storage portion 308 as needed.

[0113] In particular, the processes described above with reference to the flowcharts can be implemented as a computer software program in accordance with embodiments of the present application. For example, embodiments of the present application include a computer program product comprising a computer program carried on a computer readable medium, the computer program comprising computer programs for executing the methods illustrated by the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via the communication section 309, and / or installed from the removable medium 311. When the computer program is executed by the central processing unit (CPU) 301, various functions defined in the present application are executed.

[0114] Note that specific examples of the computer readable storage medium can include one or more of a volatile memory, a non-volatile memory, a magnetic storage medium, an optical storage medium, or a portable storage medium. In addition, the computer readable storage medium can also be any tangible medium that is suitable for storing or carrying a program. Specifically, the computer readable storage medium of the present application can be any tangible medium that can be used to store or carry the program codes of the application and that can be accessed by a general purpose or special purpose computer system.

[0115] The flowcharts and block diagrams in the drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present application. In this regard, each block in the flowcharts or block diagrams can represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions noted in the blocks can occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks can sometimes be executed in the reverse order, depending upon the functionality involved.

[0116] Specifically, the land contract right change identification system of the embodiment includes a processor and a memory, and the memory stores a computer program which, when executed by the processor, implements the land contract right change identification method based on remote sensing images provided by the above embodiment.

[0117] As another aspect, the present application also provides a computer readable storage medium, which can be included in the land contract right change identification system described in the above embodiments, or can exist independently without being assembled into the land contract right change identification system. The storage medium carries one or more computer programs, which, when executed by a processor of the land contract right change identification system, cause the land contract right change identification system to implement the land contract right change identification method based on remote sensing images provided in the above embodiments.

[0118] The above embodiments are only used to illustrate the technical solutions of the present application, but not limit the present application; although the present application has been described in detail with reference to the foregoing embodiments, those ordinarily skilled in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacements to some technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application.

[0119] In the above embodiments, according to the context, the term "when" can be interpreted as meaning "if" or "after" or "in response to determining" or "in response to detecting". Similarly, according to the context, the phrase "upon determining" or "if detecting (the stated condition or event)" can be interpreted as meaning "if determining" or "in response to determining" or "upon detecting (the stated condition or event)" or "in response to detecting (the stated condition or event)".

[0120] Those ordinarily skilled in the art can understand that all or part of the processes in the above embodiments can be instructed by a computer program to relevant hardware, and the program can be stored in a computer readable storage medium, and when executed, can include the processes of the above method embodiments. The foregoing storage medium includes ROM or random storage memory RAM, magnetic disc or optical disc, and various program code storage media.

Claims

1. A method for identifying changes in land contract rights based on remote sensing images, characterized in that, The method is applied to a land contract right change identification system, and the method comprises: Obtaining continuous high-resolution remote sensing images in a target research area, the high-resolution remote sensing images being obtained continuously at a set time interval, and the high-resolution remote sensing images including images in key farming activity periods in a crop growing season; Calculating vegetation index time series data by using the high-resolution remote sensing images; Determining a vegetation index mutation time point as a farming activity time according to a mean rate of change of the vegetation index time series data in a time dimension, and calculating a mean standard deviation of the vegetation index time series data in a spatial dimension in each period as a tillage uniformity, the vegetation index mutation time point being a time point at which a change rate of adjacent two periods of the vegetation index time series data is greater than a preset change rate threshold; Obtaining a time difference value according to an absolute difference value of adjacent plot farming activity times, obtaining a uniformity difference value according to an absolute difference value of adjacent plot tillage uniformities, and obtaining a farming activity difference index by weighting the time difference value and the uniformity difference value after normalization processing according to a preset weight; When the farming activity difference index of a target plot is greater than a preset difference threshold, marking the target plot as a suspected contract right change plot; When a correlation coefficient of a farming activity difference index time series curve of the suspected contract right change plot and a historical time series curve is less than a preset correlation threshold, determining the suspected contract right change plot as a contract right change plot.

2. The method of claim 1, wherein, The step of determining a vegetation index mutation time point as a farming activity time according to a mean rate of change of the vegetation index time series data in a time dimension specifically comprises: Dividing the vegetation index time series data into a plurality of observation periods according to a set observation period; Calculating a daily mean value of the vegetation index in each observation period; Calculating a difference between the daily mean values of the vegetation index of adjacent two observation periods to obtain a vegetation index change amount, and calculating a mean rate of change of the vegetation index time series data according to the vegetation index change amount; Extracting a time point at which the mean rate of change is greater than the preset change rate threshold as the vegetation index mutation time point, and determining a farming activity time according to a distribution feature of the vegetation index mutation time point, the farming activity time being a time point with a density value greater than a preset density threshold in a time series set of the vegetation index mutation time points.

3. The method of claim 1, wherein, The step of obtaining a farming activity difference index by weighting the time difference value and the uniformity difference value after normalization processing according to a preset weight specifically comprises: Calculating a time span of the farming activity time of each plot in the research area, and obtaining a normalized time difference value by dividing the time difference value by the time span; Calculating a change range of the tillage uniformity of each plot in the research area, and obtaining a normalized uniformity difference value by dividing the uniformity difference value by the change range; Determining a time difference weight and a uniformity difference weight based on an indication degree of farming activities to land contract right change, multiplying the normalized time difference value and the time difference weight, multiplying the normalized uniformity difference value and the uniformity difference weight, and adding the products to obtain the farming activity difference index.

4. The method of claim 1, wherein, The step of determining the suspected contract right change plot as a contract right change plot when the correlation coefficient between the agricultural activity difference index time series curve of the suspected contract right change plot and the historical time series curve is less than a preset correlation threshold value, specifically includes: Obtaining the agricultural activity difference index time series data of the suspected contract right change plot in a plurality of historical years without contract right change; Performing average processing on the agricultural activity difference index time series data of the plurality of historical years to obtain a baseline historical time series curve; Collecting the agricultural activity difference index of each agricultural activity period in the current year of the suspected contract right change plot to generate a current time series curve; Calculating the correlation coefficient between the current time series curve and the baseline historical time series curve; When the correlation coefficient is less than the preset correlation threshold value, determining the suspected contract right change plot as a contract right change plot.

5. The method of claim 1, wherein, After the step of determining the suspected contract right change plot as a contract right change plot when the correlation coefficient between the agricultural activity difference index time series curve of the suspected contract right change plot and the historical time series curve is less than a preset threshold value, the method further includes: Obtaining the agricultural activity difference index time series curve of the surrounding plots of the contract right change plot; According to the similarity of the agricultural activity difference index time series curves of the contract right change plot and the surrounding plots, the surrounding plots are divided into a synchronous change area and a non-synchronous change area; Calculating the spatial autocorrelation coefficient of the agricultural activity difference index of the plots in the synchronous change area, and determining a concentrated and contiguous area of contract right change according to the spatial autocorrelation coefficient.

6. The method of claim 1, wherein, The step of dividing the surrounding plots into a synchronous change area and a non-synchronous change area according to the similarity of the agricultural activity difference index time series curves of the contract right change plot and the surrounding plots, specifically includes: Establishing an agricultural activity difference index time series feature vector of the contract right change plot, the time series feature vector including fluctuation amplitude, periodicity and trend; Calculating the similarity coefficient between the agricultural activity difference index time series feature vector of the surrounding plots and the time series feature vector of the contract right change plot; According to the similarity coefficient, a time series feature similarity score is determined, which reflects the consistency degree of agricultural activity change; The surrounding plots with a time series feature similarity score greater than a set threshold value are divided into the synchronous change area, and the surrounding plots with a time series feature similarity score less than or equal to a set threshold value are divided into the non-synchronous change area.

7. The method of claim 1, wherein, After the step of determining the suspected contract right change plot as a contract right change plot when the correlation coefficient between the agricultural activity difference index time series curve of the suspected contract right change plot and the historical time series curve is less than a preset threshold value, the method further includes: Obtaining the on-site verification result of the contract right change plot in a historical period, and establishing a corresponding relationship between the agricultural activity difference index time series curve and the on-site verification result; According to the correspondence, a characteristic mode of agricultural activities of different types of land contract right change is identified, and a change type of a newly identified contract right change plot is determined based on the characteristic mode of agricultural activities.

8. A land contract right change identification system, characterized in that, The land contract right change identification system comprises one or more processors and a memory; the memory is coupled with the one or more processors, and the memory is configured to store computer program codes, the computer program codes comprising computer instructions, and the one or more processors are configured to invoke the computer instructions to enable the land contract right change identification system to perform the method according to any one of claims 1-7.

9. A computer-readable storage medium comprising instructions, characterized in that, The instructions, when executed on the land contract right change identification system, enable the land contract right change identification system to perform the method according to any one of claims 1-7.

10. A computer program product, characterised in that, The computer program product, when executed on the land contract right change identification system, enables the land contract right change identification system to perform the method according to any one of claims 1-7.

Images