Automatic detection method of surface ecological change information

Abstract

The application provides an automatic detection method of surface ecological change information, comprising the following steps: according to a normalized surface comprehensive ecological index, average semi-variance of each pixel in a data set at different time intervals within an observation period is calculated; a least square method is used to estimate a trend value, and a maximum value or a minimum value of the average semi-variance at different time intervals within an interannual observation period is detected. Through the technical scheme of the application, the average semi-variance is introduced to quantitatively measure the response relationship between the time change and the surface ecological change, on the basis of which, the maximum value or the minimum value of the average semi-variance at different time intervals within the interannual observation period is detected, the extraction of the regional surface ecological change information is completed, the kernel algorithm is simple and easy to realize, the change detection precision is improved, the detection content of the ecological change is widened, the change direction and the change degree can be detected, and the result data has indication and representation.

Description

Technical Field

[0001] This invention relates to the field of ecological change assessment technology, and more specifically, to an automatic detection method for surface ecological change information. Background Technology

[0002] The process of change in the Earth's surface ecological environment is a continuous, cumulative, and ongoing process. Ecological change involves both qualitative and quantitative changes. Generally speaking, the dynamic monitoring of surface habitats can be divided into seasonal changes within the year (ecological status), abrupt changes between years (vegetation type conversion), and gradual changes (ecological degradation and improvement).

[0003] Continuous Change Detection and Classification (CCDC) is a time-series change detection algorithm that uses a simple sinusoidal model to simulate seasonal changes in land cover. It requires a sufficient number of observations for model initialization. A pixel is identified as having a land cover change when the difference between the observed image observation and the model's predicted image observation exceeds three times the root mean square error (RMSE). The model simulates the time series of each pixel's bands using Ordinary Least Squares (OLS), including interannual, inter-annual, and abrupt changes, detecting potential land cover change areas each year. However, the CCDC algorithm's simulation of pixel time series is susceptible to noise (cloud observations) and parameter settings, and the detected change areas may contain "spurious changes."

[0004] The Coupled Iterative Seasonal and Trend (BFAST) model is a time-series change detection algorithm that can detect breakpoints. It decomposes remote sensing time series into a trend term (including abrupt changes and the fitted trend between them), a seasonal term, and a noise term. Change detection is performed on monthly Landsat NDVI time series, yielding nine seasonal parameters and two trend parameters. These parameters are then used as input to a random forest classifier to generate land cover type maps for any given time period, improving the accuracy of land classification.

[0005] Object-based change detection (OBCD) is a change detection method that integrates spatiotemporal information. OBCD research focuses not on single pixel attributes, but on clusters of pixels with shared spectral physical properties at a regional scale.

[0006] The LandTrendr algorithm is a method for identifying pixel values ​​that change over time. This model runs on the Google Earth Engine (GEE) platform. Developed for time series data from multi-band satellites, it detects changes and classifies land cover before and after changes occur. The tool works with data from supported sensors and can be used to detect changes within single-band rasters. Pixel information from a single spectral band or spectral index is used to extract a trajectory for each pixel over time. The trajectory is then segmented to capture and model time periods of no change, change, and recovery from change.

[0007] In summary, it is evident that most change detection methods inevitably consider water bodies, especially large areas of constant water, which significantly impact the accuracy of the final detection results. Researchers tend to focus more on improving the accuracy of change detection, neglecting another crucial criterion for algorithm evaluation: simplicity, ease of use, and universality. Existing change detection methods suffer from problems such as complex computational processes, low computational accuracy, and limited information extraction. Summary of the Invention

[0008] The present invention aims to solve at least one of the technical problems existing in the prior art or related art.

[0009] Therefore, the purpose of this invention is to provide an automatic detection method for surface ecological change information. It introduces the mean semivariance, establishing a functional relationship between the variance of the comprehensive surface ecological index at any two time points within the observation period and the time interval between these two periods. This yields the mean semivariance of surface ecological change at the same time interval within the observation period, quantitatively measuring the response relationship between temporal change and surface ecological change. Based on this, extreme value detection identifies the maximum or minimum value of the mean semivariance at different time intervals within the interannual observation period, completing the extraction of regional surface ecological change information. The core algorithm is simple and easier to implement, improving the accuracy of change detection, broadening the scope of ecological change detection, and enabling the detection of both the direction and degree of change. Furthermore, the resulting data is indicative and representative.

[0010] To achieve the above objectives, the present invention provides an automatic detection method for surface ecological change information, comprising the following steps: Based on the normalized integrated ecological index of the land surface, the mean semivariance of each pixel in the dataset is calculated traversally at different time intervals in the dataset during the observation period. The least squares method is used to estimate the trend value, and the extreme value detection method is used to find the maximum or minimum value of the average semivariance at different time intervals within the interannual observation period.

[0011] This technical solution provides a high-density time-series change detection algorithm. It establishes a functional relationship between the variance of the comprehensive surface ecological index at any two time points within the observation period and the time interval between these two periods. This yields the average semivariogram of surface ecological changes within the same time interval during the observation period, quantitatively measuring the response relationship between temporal changes and surface ecological changes. Based on this, extreme value detection identifies the maximum or minimum value of the average semivariogram at different time intervals within the interannual observation period, completing the extraction of regional surface ecological change information. This effectively solves the problems of complex calculation processes and low change detection accuracy, achieving automatic detection and extraction of the comprehensive surface ecological index of the observation area, quantitatively measuring the intra-annual rating and interannual change direction and degree of surface ecology, eliminating the influence of human factors, and simplifying the calculation process. Furthermore, experimental verification shows that the results are strongly correlated with the surface vegetation cover type and interannual transition type of the study area, demonstrating indicative and representative characteristics.

[0012] In the above technical solution, based on the normalized comprehensive ecological index of the land surface, the average semivariance of each pixel in the dataset is calculated iteratively at different time intervals within the dataset during the observation period. Specifically, this includes the following steps: Based on the normalized comprehensive ecological index of the land surface, and following the first preset formula, the average semivariance of each pixel at different time intervals in the dataset during the observation period is calculated iteratively. The first preset formula is , Where hΔt represents the time interval; x i The i-th comprehensive ecological index of the land surface is represented by the value of h, which takes the values ​​1, 2, ... (n-1), where n represents the number of comprehensive ecological indices of the land surface in the data during the observation period; x (i +hΔt) Characterized as with x i The next comprehensive ecological index of the land surface is given by the time interval hΔt; N hΔt S is the number of integrated ecological indices of the land surface, represented by the time interval hΔt. hΔt It is represented as the average semivariance over the dataset with time interval hΔt.

[0013] In this technical solution, a method for calculating the mean semivariance is constructed. By introducing the mean semivariance, it is possible to quantitatively measure the response relationship between temporal changes and changes in the land surface ecology.

[0014] Specifically, an algorithm is written on the ArcGIS-Python language software platform to calculate the average semivariance of a cell. The number of iterations is determined by the actual number of years to be calculated, thereby obtaining all the average semivariances of the cell at different time intervals.

[0015] In any of the above technical solutions, preferably, the least squares method is used to estimate the trend value, and the maximum or minimum value of the average semivariance under different time intervals within the interannual observation period is detected by extreme value detection, specifically including the following steps: According to the second preset formula, the trend value is estimated using the least squares method. The second preset formula is , Where k represents the trend value of the comprehensive surface ecological index during the observation period, and t i Characterized by the year of observation i, y i It is represented by the comprehensive ecological index of the land surface calculated for the corresponding year i. Characterized as the average value over the years of observation. Characterized by the average value of the comprehensive surface ecological index during the observation period, If k > 0, then the maximum value of the mean semivariogram is determined; if k < 0, then the minimum value of the mean semivariogram is determined.

[0016] In this technical solution, the least squares method is used to estimate the trend value k. When k>0, the maximum value of the mean semivariogram is determined, that is, the maximum value is calculated. When k<0, the minimum value of the mean semivariogram is determined, that is, the minimum value is calculated. Thus, the correlation between variables can be effectively described, the range of wetland ecological change and the main wetland cover types of the change can be quickly determined, and the significance of wetland ecological change (intra-annual and inter-annual) can be quantitatively measured.

[0017] In the above technical solution, preferably, the automatic detection method for surface ecological change information further includes the following steps: saving the maximum or minimum average semivariance value obtained for each pixel; assigning projection parameters consistent with the original data map to the result data, saving the data, and completing the extraction of surface ecological change information.

[0018] In the above technical solution, preferably, the automatic detection method for surface ecological change information further includes the following steps: An annual ecological rating scheme for the land surface is pre-constructed, a comprehensive ecological index for the land surface is calculated, and the ecological status of the land surface is rated during the observation period.

[0019] In this technical solution, by pre-constructing an annual ecological rating scheme for the land surface, calculating a comprehensive ecological index for the land surface, and rating the ecological status of the land surface during the observation period, data support is provided for the calculation of the average semivariance of pixels.

[0020] It should be noted that the comprehensive surface ecological index is an integrated value obtained from indicators such as vegetation, water bodies, and soil through principal component analysis, used to evaluate the ecological quality of the monitored area. Typically, the comprehensive surface ecological index is converted into an 85-point scale or a percentage scale, and further divided into five equally spaced value gradients, corresponding to five ecological status levels: poor, relatively poor, moderate, good, and excellent.

[0021] In the above technical solution, preferably, the pre-construction of the annual ecological rating scheme for the land surface, the calculation of the comprehensive ecological index of the land surface, and the rating of the land surface ecological status within the observation period specifically include the following steps: using remote sensing image data collected from most parts of China during the growing season from April to October, an annual time-series dataset of ecological factors is constructed, including the land surface vegetation index, water body index, soil surface temperature, and land surface reflectance; based on the annual time-series dataset, a water body mask is applied for the corresponding year to remove large areas of water bodies; principal component analysis is used to determine the weight of each ecological factor and calculate the comprehensive ecological index of the land surface in the observation area; based on the annual comprehensive ecological index of the land surface in the observation area, regional cropping and ecological classification are performed to obtain the land surface ecological status rating.

[0022] In this technical solution, based on the time series data set of the Land Resources Satellite, the change information of ecological factors such as hydrology, vegetation and soil in the study area is calculated and extracted. The calculation results of the maximum difference of ecological factors are subjected to layer stacking and principal component analysis to obtain the first principal component (PCA-1). The first principal component is used to classify the significance of the data and obtain the intra-annual / inter-annual evaluation results of the surface ecological changes in the study area.

[0023] In the above technical solution, preferably, an annual time-series dataset of ecological factors is constructed using remote sensing image data collected from April to October in most parts of China during the growing season, specifically including the following steps: Using the actual collected time-series data of remote sensing images for j=4, 5, 6, 7, 8, 9, and 10 months, and selecting image data with cloud cover <10%, radiometric calibration and atmospheric correction of the visible light band of the satellite images were performed using the remote sensing software ERDAS platform. Normalized surface vegetation index, normalized water index, soil surface temperature, and surface reflectance were obtained through band calculations. Perform outlier detection and interpolate outliers using a filtering algorithm; Extreme value calculations are used to determine the surface ecological factors during the growing season in the observation area, thus obtaining the annual ecological status of the observation area. Statistical analysis was conducted on the annual ecological factor range of the observation area to determine the normalized range of annual soil surface temperature and surface reflectance, thus obtaining the normalized soil surface temperature and normalized surface reflectance. Imputing missing values ​​in normalized data using low-value imputation; Repeat the above steps to construct a time-series dataset of surface ecological factors in the annual observation area over many years.

[0024] In this technical solution, remote sensing satellite data is preprocessed through noise removal, interpolation, and data normalization (or relative radiometric correction) to achieve comparability of different ecological indicators. Remote sensing time series reconstruction is a prerequisite for change detection.

[0025] It should be noted that when using low-value imputation to fill in missing values ​​in normalized data, the value "0" is generally used.

[0026] In the above technical solution, preferably, the direction of surface ecological change and land transformation type can be obtained based on the extracted surface ecological change information.

[0027] The present invention provides a computer device, which includes a processor. The processor executes a computer program stored in a memory to implement the steps of the automatic detection method for surface ecological change information as proposed in the present invention.

[0028] All the beneficial effects of the automatic detection method for surface ecological change information proposed by the technical solution of the present invention will not be repeated here.

[0029] The present invention provides a computer-readable storage medium storing a computer program thereon. When the computer program is executed by a processor, it implements the steps of the automatic detection method for surface ecological change information proposed in the present invention.

[0030] All the beneficial effects of the automatic detection method for surface ecological change information proposed by the technical solution of the present invention will not be repeated here.

[0031] The automatic detection method for surface ecological change information proposed in this invention has the following beneficial technical effects: (1) The present invention proposes an automatic detection method for surface ecological change information. It introduces the average semivariance and establishes a functional relationship between the variance of the comprehensive ecological index of the surface at any two time points within the observation period and the time interval between the two periods of data. The average semivariance of surface ecological change under the same time interval within the observation period can be obtained, which can quantitatively measure the response relationship between time change and surface ecological change.

[0032] (2) The automatic detection method for surface ecological change information proposed in this invention, based on the introduction of average semivariance, detects the maximum or minimum value of average semivariance at different time intervals within the interannual observation period, thereby extracting regional surface ecological change information. The kernel algorithm is simple and easier to implement, which improves the accuracy of change detection, broadens the detection content of ecological change, and can detect the direction and degree of change. Moreover, the result data is indicative and representative.

[0033] (3) The automatic detection method for surface ecological change information proposed in this invention effectively solves the problems of complex calculation process and low change detection accuracy. It realizes automatic detection and extraction of the comprehensive ecological index of the surface in the observation area, quantitatively measures the intra-annual rating and inter-annual change direction and degree of surface ecology, eliminates the influence of human factors, and simplifies the calculation process. Experiments have shown that the results are strongly correlated with the surface vegetation cover type and inter-annual transition type in the study area, and have indicative and representative properties.

[0034] Additional aspects and advantages of the invention will be set forth in the description which follows, and in part will be obvious from the description or may be learned by practice of the invention. Attached Figure Description

[0035] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which: Figure 1 A flowchart illustrating an automatic detection method for surface ecological change information according to an embodiment of the present invention is shown. Figure 2 A flowchart illustrating the process of rating the surface ecological status during the observation period is shown. Figure 3 The map shows the surface ecological status rating results for the observation period; Figure 4 A schematic diagram illustrating the design principle of the semivariance algorithm is shown. Figure 5 A trend analysis diagram of surface time-series data transformation is shown; Figure 6 The diagram shows a comparison of experimental data verification points and ecological change detection results obtained using the automatic detection method for surface ecological change information proposed in this invention. Detailed Implementation

[0036] To better understand the above-mentioned objectives, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the embodiments and features of this application can be combined with each other.

[0037] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and therefore the scope of protection of the invention is not limited to the specific embodiments disclosed below.

[0038] like Figure 1 As shown, an automatic detection method for surface ecological change information according to an embodiment of the present invention includes the following steps: S102, based on the normalized integrated ecological index of the land surface, calculate the average semivariance of each pixel in the dataset at different time intervals in the dataset during the observation period; S104 uses the least squares method to estimate trend values ​​and extreme value detection to find the maximum or minimum value of the average semivariance at different time intervals within the interannual observation period.

[0039] This embodiment provides a high-density time series change detection algorithm. It establishes a functional relationship between the variance of the comprehensive surface ecological index at any two time points within the observation period and the time interval between these two periods. This yields the average semivariogram of surface ecological changes within the same time interval during the observation period, quantitatively measuring the response relationship between temporal changes and surface ecological changes. Based on this, extreme value detection identifies the maximum or minimum value of the average semivariogram at different time intervals within the interannual observation period, completing the extraction of regional surface ecological change information. This effectively solves the problems of complex calculation processes and low change detection accuracy, achieving automatic detection and extraction of the comprehensive surface ecological index of the observation area, quantitatively measuring the intra-annual rating and interannual change direction and degree of surface ecology, eliminating the influence of human factors, and simplifying the calculation process. Furthermore, experimental verification shows that the results are strongly correlated with the surface vegetation cover type and interannual transition type of the study area, demonstrating indicative and representative characteristics.

[0040] An automatic detection method for surface ecological change information according to an embodiment of the present invention, using ArcGIS 10.8 as the software platform and Python 3.5 as the algorithm programming language, includes the following steps: S202, Collection of Time Series Data The actual time-series data collected in this invention consisted of remote sensing imagery data for April, May, June, July, August, September, and October (the approximate range of the growing season in northern China is April to October). Imagery with cloud cover <10% was selected. Based on the maximum value and maximum variance of the time-series data set within the year, the optimal state of the surface ecosystem and its stability during the growing season can be quantified, ultimately completing the surface ecosystem rating for the growing season. The experimental data are shown in Table 1 below.

[0041] S204 Data Preprocessing Radiometric calibration and atmospheric correction of satellite imagery in the visible light band were performed using the remote sensing software ERDAS platform. Atmospheric correction can be achieved using the FLAASH model to correct the atmospheric conditions of the calibration data. Normalized surface vegetation index (NDVI), normalized water index (NDWI), soil surface temperature (LST), and surface reflectance (Albedo) were obtained through band calculations. These ecological factors represent the productivity of surface ecology, soil moisture content, and surface exposure, respectively. Outlier detection is performed, and outlier interpolation is used to fill in the gaps through a filtering algorithm. Specifically, the filtering algorithm statement is: Con(IsNull("raster"), FocalStatistics("raster", NbrRectangle(5, 5, "CELL"), "MEAN"),"raster"); The comparability of different ecological indicators is achieved by normalizing ecological factors. The range of ecological factor values ​​in the annual growing season of the observation area is statistically analyzed to determine the normalized range of soil surface temperature and surface reflectance in the annual growing season, and the normalized soil surface temperature and normalized surface reflectance are obtained. Finally, the normalization (0-1) is completed. After normalization, the data is filled with null values. The null values ​​to be filled are determined. The null values ​​to be filled are determined by filling the null values ​​with the lowest value - 0. The algorithm statement is: Con(IsNull("raster"),0,"raster"); Repeat the above steps to obtain a time-series dataset of surface ecological factors in the annual observation area over many years.

[0042] The main algorithm design for remote sensing time series reconstruction involves numerous preprocessing issues related to remote sensing satellite data, such as noise removal, interpolation imputation, and data normalization (or relative radiometric correction). These issues can be collectively referred to as the technical problems in the remote sensing time series reconstruction process. Remote sensing time series reconstruction is a prerequisite for change detection.

[0043] Calculation of S206 Comprehensive Ecological Index of Land Surface The removal of large areas of water eliminates the impact of water bodies on ecological rating results and improves the detection accuracy of surface ecological changes. The water mask is created using the following statement: outRas = Con(outNDWI>= 0.27, 0,Con(outNDWI<0.27,1,0)). Principal component analysis (PCA) calculates the annual comprehensive ecological index of the land surface in the observation area and assigns weights to various ecological factors, replacing the influence of human assignment on the final calculation results. Complete the regional scope trimming and ecological classification to obtain regional ecological rating results; Specific technical processes are as follows Figure 2 As shown, the surface ecological status rating results during the observation period are as follows: Figure 3 As shown.

[0044] S208 Mean Semivariance Calculation and Extreme Value Extraction Based on the regional ecological rating results of the annual comprehensive surface ecological index, an algorithm was developed on the ArcGIS-Python software platform to calculate the actual mean semivariance of pixels. The number of iterations was determined according to the actual number of years to be calculated, resulting in the average semivariance of all pixels at different time intervals. The design principle of the semivariance algorithm is as follows: Figure 4 As shown, the solid circles represent the time points (T) when the remote sensing data was acquired. The average semivariance is calculated on the dataset at time intervals Δt, 2Δt, 3Δt...hΔt (the value of h depends on the actual amount of data used for calculation). The formula is as follows: , Where hΔt represents the time interval; x i The i-th comprehensive ecological index of the land surface is represented by the value of h, which takes the values ​​1, 2, ... (n-1), where n represents the number of comprehensive ecological indices of the land surface in the data during the observation period; x (i +hΔt) Characterized as with x i The next comprehensive ecological index of the land surface is given by the time interval hΔt; N hΔt S is the number of integrated ecological indices of the land surface, represented by the time interval hΔt. hΔt It is represented as the average semivariance over a dataset with time intervals hΔt. This function can effectively describe the correlation between variables, thereby enabling rapid determination of the extent of wetland ecological change and the main wetland cover types involved in the change, and quantitatively measuring the significance of wetland ecological change (intra-annual / inter-annual).

[0045] Trend analysis. The least squares method is used to estimate the trend value. If k>0, the mean semivariance is taken as MAX() to calculate the maximum value; if k<0, the mean semivariance is taken as MIN() to calculate the minimum value.

[0046] The least squares method is used to estimate trend values, and the equation calculation formula is as follows: , Where k represents the trend value of the comprehensive surface ecological index during the observation period. t i Characterized by the year of observation i, y i It is represented by the comprehensive ecological index of the land surface calculated for the corresponding year i. Characterized as the average value over the years of observation. It is represented by the average value of the comprehensive ecological index of the land surface.

[0047] The surface time series data transformation trend analysis diagram is shown below. Figure 5 As shown.

[0048] The maximum or minimum value of the average semivariance is obtained by saving each pixel.

[0049] S210, complete the extraction of regional surface ecological change information. Determine whether to end the current calculation based on the initially set maximum number of years. Assign projection parameters to the resulting data, ensuring that the projection parameters are consistent with the original data map parameters; Finally, the data is saved, completing the extraction of information on changes in the Earth's surface ecosystem.

[0050] A comparison diagram of experimental data verification points and ecological change detection results obtained using the automatic detection method for surface ecological change information proposed in this invention is shown below. Figure 6 As shown in the table, the experimental data exhibits a strong correlation with the surface vegetation cover type and interannual transition type in the study area, demonstrating indicative and representative characteristics. The results of the surface ecological change detection accuracy analysis of the experimental data are shown in Table 2 below.

[0051] In summary, the proposed automatic detection method for surface ecological change information effectively solves the problems of complex calculation processes and low detection accuracy. It achieves automatic detection and extraction of the comprehensive surface ecological index of the observation area, quantitatively measures the intra-annual rating and inter-annual change direction and degree of surface ecology, eliminates the influence of human factors, and simplifies the calculation process. Experiments show that the results are strongly correlated with the surface vegetation cover type and inter-annual transition type of the study area, demonstrating indicative and representative characteristics.

[0052] The steps in the method of this invention can be adjusted, combined, or deleted according to actual needs.

[0053] The units in the device of the present invention can be merged, divided, or deleted according to actual needs.

[0054] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, including read-only memory (ROM), random access memory (RAM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), one-time programmable read-only memory (OTPROM), electrically-Erasable Programmable Read-Only Memory (EEPROM), compactdisc read-only memory (CD-ROM) or other optical disc storage, disk storage, magnetic tape storage, or any other computer-readable medium capable of carrying or storing data.

[0055] In the description of this specification, the terms "one embodiment," "some embodiments," "specific embodiment," etc., refer to a specific feature, structure, material, or characteristic described in connection with that embodiment or example, which is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0056] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. An automatic detection method for information on changes in land surface ecology, characterized in that, Includes the following steps: S1, pre-construct an annual ecological rating scheme for the land surface, calculate the comprehensive ecological index of the land surface, and conduct a rating of the land surface ecological status during the observation period; S2, based on the normalized integrated ecological index of the land surface, calculates the average semivariance of each pixel in the dataset at different time intervals during the observation period; S3 uses the least squares method to estimate trend values ​​and extreme value detection to find the maximum or minimum value of the average semivariance at different time intervals within the interannual observation period; S1 specifically includes the following steps: Using remote sensing image data collected from April to October of the growing season in the region, an annual time-series dataset of ecological factors was constructed, which includes surface vegetation index, water index, soil surface temperature and surface reflectance. Based on the annual time series dataset, perform water body masking for the corresponding year to remove large areas of water. Principal component analysis was used to determine the weights of each ecological factor and to calculate the comprehensive ecological index of the land surface in the observation area. Based on the comprehensive ecological index of the observed area, regional cropping and ecological classification are performed to obtain a surface ecological status rating; S2 specifically includes the following steps: Based on the normalized comprehensive ecological index of the land surface, and following the first preset formula, the average semivariance of each pixel in the dataset at different time intervals within the observation period is calculated iteratively. The first preset formula is ， Where hΔt represents the time interval; x i The i-th comprehensive ecological index of the land surface is represented by the value of h, which takes the values ​​1, 2, ... (n-1), where n represents the number of comprehensive ecological indices of the land surface in the data during the observation period; x (i +hΔt) Characterized as with x i The next comprehensive ecological index of the land surface is given by the time interval hΔt; N hΔt S is the number of integrated ecological indices of the land surface, represented by the time interval hΔt. hΔt It is represented as the average semivariance over the dataset with time interval hΔt; S3 specifically includes the following steps: According to the second preset formula, the trend value is estimated using the least squares method. The second preset formula is ， Where k represents the trend value of the comprehensive surface ecological index during the observation period, and t i Characterized by the year of observation i, y i It is represented by the comprehensive ecological index of the land surface calculated for the corresponding year i. Characterized as the average value over the years of observation. Characterized by the average value of the comprehensive surface ecological index during the observation period, If k > 0, then the maximum value of the mean semivariogram is determined; if k < 0, then the minimum value of the mean semivariogram is determined.

2. The automatic detection method for surface ecological change information according to claim 1, characterized in that, It also includes the following steps: S4, the maximum or minimum value of the average semivariance obtained by saving each pixel; S5 assigns projection parameters consistent with the original data map to the resulting data, saves the data, and completes the extraction of information on surface ecological changes.

3. The automatic detection method for surface ecological change information according to claim 1, characterized in that, Using remote sensing imagery data collected from the region during the growing season (April to October), an annual time-series dataset of ecological factors was constructed, including the following steps: Using the actual collected time-series remote sensing image data for months j=4, 5, 6, 7, 8, 9, and 10, and selecting image data with cloud cover <10%, radiometric calibration and atmospheric correction of the visible light band of the satellite imagery were performed using the remote sensing software ERDAS platform. Normalized surface vegetation index, normalized water index, soil surface temperature, and surface reflectance were obtained through band calculations. Perform outlier detection and interpolate outliers using a filtering algorithm; Extreme value calculations are used to determine the surface ecological factors during the growing season in the observation area, thus obtaining the annual ecological status of the observation area. Statistical analysis was conducted on the annual ecological factor range of the observation area to determine the normalized range of annual soil surface temperature and surface reflectance, thus obtaining the normalized soil surface temperature and normalized surface reflectance. Imputing missing values ​​in normalized data using low-value imputation; Repeat the above steps to construct a time-series dataset of surface ecological factors in the annual observation area over many years.

4. The automatic detection method for surface ecological change information according to claim 2, characterized in that, Based on the extracted information on surface ecological changes, the direction of surface ecological changes and the type of land transformation are obtained.

5. A computer device, characterized in that, The computer device includes a processor, which executes a computer program stored in a memory to implement the automatic detection method for surface ecological change information as described in any one of claims 1 to 3.

6. A computer-readable storage medium, characterized in that, It stores a computer program, which, when executed by a processor, implements the automatic detection method for surface ecological change information as described in any one of claims 1 to 3.

Images