Method, device and equipment for identifying full life cycle of abandoned farmland

Abstract

The invention provides a method, a device and equipment for identifying a full life cycle of abandoned farmland. The method comprises the following steps: acquiring a surface vegetation end member time sequence image and a surface soil end member time sequence image of a target area; respectively analyzing the earth surface vegetation end member time sequence image and the earth surface soil end member time sequence image to obtain an optimal fitting curve and corresponding change characteristic parameters of the earth surface vegetation end member time sequence and the earth surface soil end member time sequence; and determining the abandoned farmland type of the target area based on the optimal fitting curve and the corresponding change characteristic parameters. According to the invention, the vegetation-soil end member time sequence representing the abandoned land surface process of the cultivated land is utilized to carry out cultivated land abandoned full life cycle modeling and cultivated land abandoned scene identification integrated with multi-dimensional end member time sequence interaction information, so that an efficient and high-precision cultivated land abandoned full life cycle monitoring method is formed; technical support is provided for intelligent monitoring and supervision of natural resources such as non-agrochemical resources and non-grain resources of cultivated land.

Description

Technical Field

[0001] The present invention relates to the technical field of ecological environment monitoring, and in particular to a method, device and equipment for identifying abandoned farmland throughout its life cycle. Background Art

[0002] Abandoned farmland refers to the state of idle or underutilized farmland caused by the subjective reasons of producers and operators who give up farming. Scientific and effective monitoring of abandoned farmland is an important basis for ensuring the sustainable use of natural resources and food security. At present, the monitoring of abandoned farmland based on the classification results of land use and cover types for many years is one of the mainstream methods. It mainly uses the classification results of land use and cover types year by year to carry out rule-based judgment of abandoned farmland. In this process, it is necessary not only to analyze the land use and cover types, but also to identify the abandoned farmland areas based on the duration of non-cultivated land status as the judgment standard. This method is difficult to avoid the error transmission caused by the error of land use and cover classification results, resulting in a high error in the identification of abandoned farmland, and lacks an understanding of the entire life cycle of abandoned farmland. In addition, the change vector analysis based on the spectral characteristics of the images before and after or the vegetation index has also been used for the identification of abandoned farmland, but this method is difficult to describe the process of abandoned farmland, and it is difficult to fully grasp the actual path of the conversion of abandoned state, resulting in a high uncertainty problem in the identification results.

[0003] With the continuous development of time series analysis methods, some time series analysis methods based on vegetation indices have also been used to identify abandoned farmland. However, they often extract some local features of the time series and then use machine learning and other methods to identify abandoned farmland. There is a lack of monitoring of the entire life cycle of abandoned farmland. At the same time, the information contained in the vegetation index is limited, and the coordinated changes of other elements such as soil are not considered. It is difficult to fully explore the interactive information of the landscape elements of abandoned farmland, which limits the accuracy and efficiency of the identification of abandoned farmland scenes. Summary of the invention

[0004] To this end, the present invention provides a method, device and equipment for monitoring abandoned farmland throughout its life cycle, aiming to solve the technical problems in the prior art regarding the accuracy and efficiency of identifying abandoned farmland scenarios.

[0005] To achieve the above objectives, the present invention adopts the following technical solutions:

[0006] According to a first aspect of the present invention, the present invention provides a method for monitoring the whole life cycle of abandoned farmland, the method comprising:

[0007] Obtain end-member time series images of surface vegetation and end-member time series images of surface soil in the target area;

[0008] Analyzing the surface vegetation endmember time series image and the surface soil endmember time series image respectively to obtain the optimal fitting curves of the surface vegetation endmember time series and the surface soil endmember time series and the corresponding change characteristic parameters;

[0009] Based on the optimal fitting curve of the target area and the corresponding change characteristic parameters, the type of abandoned farmland in the target area is determined.

[0010] Furthermore, the surface vegetation endmember time series image and the surface soil endmember time series image are analyzed respectively to obtain the optimal fitting curves of the surface vegetation endmember time series and the surface soil endmember time series and the corresponding change characteristic parameters, including:

[0011] Adaptively fitting the surface vegetation endmember time series image and the surface soil endmember time series image to obtain the optimal fitting curve corresponding to the surface vegetation endmember time series and the surface soil endmember time series;

[0012] The characteristic parameters of the best fitting curve are determined by using a parameter selection rule that matches the best fitting curve as the change characteristic parameters.

[0013] Furthermore, the adaptive fitting is performed on the surface vegetation endmember time series image and the surface soil endmember time series image to obtain the optimal fitting curve corresponding to the surface vegetation endmember time series and the surface soil endmember time series, including:

[0014] Setting the forms of linear fitting, logical fitting and double logical fitting to respectively fit the surface vegetation end-member time series image and the surface soil end-member time series image to obtain a linear fitting curve, a logical fitting curve and a double logical fitting curve;

[0015] The residual root mean square error of each fitting curve is calculated, and the best fitting curve is determined based on the calculation results.

[0016] Furthermore, the calculating the residual root mean square error of each fitting curve and determining the best fitting curve based on the calculation result includes:

[0017]

[0018] Model optimal =argmin(RMSE Linear ,RMSE Logistic ,RMSE DoubleLogistic )

[0019] Among them, RMSE Linear is the residual root mean square error of the linear fitting curve; RMSE Logisticis the residual root mean square error of the logistic fitting curve; RMSE DoubleLogistic is the residual root mean square error of the double logistic fitting curve; Model optimal is the best fitting curve; β0 and β1 are the fitting parameters of the linear fitting curve; L, k, and t0 are the fitting parameters of the logistic fitting curve; L1, L2, k1, k2, t 0,1 and t 0,2 are the fitting parameters of the double logic fitting curve; i and x i are vegetation / soil endmember time series values ​​and years, respectively.

[0020] Furthermore, the method of determining the characteristic parameters of the best fitting curve using the parameter selection rule matching the best fitting curve as the change characteristic parameters includes:

[0021] If the optimal fitting curve is a linear fitting curve, the linear trend of the linear fitting curve is used as the change characteristic parameter;

[0022] If the optimal fitting curve is a logic fitting curve, the turning start time, turning end time and curve change amplitude in the logic fitting curve are used as the change characteristic parameters;

[0023] If the optimal fitting curve is a dual logic fitting curve, the turning start time, the second turning time, the peak / valley arrival time, the peak / valley duration time and the curve change amplitude in the dual logic fitting curve are used as the change characteristic parameters.

[0024] Further, the change characteristic parameters of the logic fitting curve and the dual logic fitting curve are determined by second-order derivatives;

[0025] The formula of the second-order derivative is expressed as follows:

[0026]

[0027] Where z = e at-b ; a and b are fitting parameters of the logic fitting curve respectively; the turning start time t0 of the logic fitting curve is the maximum value of the second-order derivative, and the turning end time t1 is the minimum value of the second-order derivative; the turning start time t0 of the double logic fitting curve is the maximum value of the second-order derivative of the first logic fitting curve, the peak / valley value arrival time t1 is the minimum value of the second-order derivative of the first logic fitting curve, and the turning time t2 again is the maximum value of the second-order derivative of the second logic fitting curve;

[0028] and / or,

[0029] The formulas for the curve variation ranges of the logic fitting curve and the dual logic fitting curve are expressed as follows:

[0030] Mag=f(t1)-f(t0)

[0031] and / or,

[0032] The formula for the peak / valley duration of the dual logic fitting curve is expressed as follows:

[0033] Dur=t2-t1

[0034] Among them, t0 is the starting time of the turning point; t1 is the ending time of the turning point or the peak / valley value arrival time; t2 is the next turning point time.

[0035] Furthermore, the determining of the type of abandoned farmland in the target area based on the optimal fitting curve and the corresponding change characteristic parameters of the target area includes:

[0036] If the optimal fitting curve of the target area is a logical fitting curve, the type of abandoned farmland in the target area is determined using the following formula:

[0037]

[0038] Among them, K is a positive empirical threshold; 1 represents that the target area is an abandoned farmland area; 0 represents that the target area is an abandoned non-farmland area; Mag v is the change range of the vegetation logic fitting curve in the target area; Mag s is the change range of the soil logic fitting curve of the target area;

[0039] and / or,

[0040] If the optimal fitting curve of the target area is a dual logic fitting curve, the type of abandoned farmland in the target area is determined using the following formula:

[0041]

[0042] Among them, Mag v is the variation range of the vegetation double logic fitting curve in the target area; Mag s is the change amplitude of the soil double logic fitting curve of the target area; 1 represents that the target area is a potential cultivated land change area; 0 represents that the target area is a non-cultivated land abandonment area;

[0043] When the target area is a potential cultivated land change area, the type of abandoned cultivated land in the target area is determined using the following formula:

[0044]

[0045] Among them, Dur vDur is the valley duration of the vegetation double logic fitting curve in the target area; s is the peak duration of the soil double logic fitting curve of the target area; T is the preset duration; 1 represents that the target area is a potential abandoned area; 0 represents that the target area is a fallow area.

[0046] Furthermore, the type of abandoned farmland in the target area in the later period is determined based on the land use type, and the formula is as follows:

[0047]

[0048] Among them, LU later is the later land use type of the target area; Non_AG ​​is non-cultivated land; AG is cultivated land; 1 represents that the target area is an abandoned area; 0 represents that the target area is a reclaimed area after abandonment.

[0049] According to a second aspect of the present invention, the present invention provides a device for identifying abandoned farmland throughout its life cycle, the device comprising:

[0050] A data acquisition module is used to acquire end-member time series images of surface vegetation and end-member time series images of surface soil in a target area;

[0051] A data analysis module, used to analyze the surface vegetation end-member time series image and the surface soil end-member time series image respectively, to obtain the optimal fitting curves of the surface vegetation end-member time series and the surface soil end-member time series and the corresponding change characteristic parameters;

[0052] The type identification module is used to determine the type of abandoned farmland in the target area based on the optimal fitting curve of the target area and the corresponding change characteristic parameters.

[0053] According to the third aspect of the present invention, the present invention provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that when the processor executes the computer program, the method for identifying abandoned farmland throughout its life cycle as described in any one of the first aspect of the present invention is implemented.

[0054] The present invention adopts the above technical solution and has at least the following beneficial effects:

[0055] Through the scheme of the present invention, the surface vegetation endmember time series images and the surface soil endmember time series images of the target area are obtained; the surface vegetation endmember time series images and the surface soil endmember time series images are analyzed respectively to obtain the optimal fitting curves and corresponding change characteristic parameters of the surface vegetation endmember time series and the surface soil endmember time series; based on the optimal fitting curves and corresponding change characteristic parameters of the target area, the type of abandoned farmland in the target area is determined. Therefore, by using the vegetation-soil endmember time series data that characterizes the surface process of abandoned farmland, the whole life cycle modeling of abandoned farmland and the identification of abandoned farmland scenes that integrate multi-dimensional endmember time series interactive information are carried out, forming an efficient and high-precision monitoring method for the whole life cycle of abandoned farmland, providing technical support for the intelligent monitoring and supervision of natural resources such as "non-agriculturalization" and "non-grainization" of cultivated land.

[0056] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0058] Figure 1 A schematic diagram of a process of identifying a method for abandoned farmland throughout its life cycle provided by an embodiment of the present invention is shown;

[0059] Figure 2 A brief schematic diagram showing three types of optimal fitting curves provided by an embodiment of the present invention;

[0060] Figure 3 A brief schematic diagram of a changing characteristic parameter provided by an embodiment of the present invention is shown;

[0061] Figure 4 A simplified schematic diagram of calculating the first-order derivative and the second-order derivative of a characteristic parameter of a change provided by an embodiment of the present invention is shown;

[0062] Figure 5 A schematic diagram of the structure of a device for identifying abandoned farmland throughout its life cycle provided by an embodiment of the present invention is shown;

[0063] Figure 6 A schematic diagram of the physical structure of a computer device provided by an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0064] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0065] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the statement "include..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0066] The embodiment of the present invention provides a method for identifying abandoned farmland throughout its life cycle. Figure 1 As shown, at least the following steps S101 to S103 may be included:

[0067] Step S101, obtaining end-member time series images of surface vegetation and end-member time series images of surface soil in a target area.

[0068] Different from the traditional vegetation index, the end member can represent the proportion of different objects in the mixed pixel in a physically meaningful way, which is helpful for extracting multi-dimensional surface element information. Therefore, the embodiment of the present invention obtains the end member time series image based on the surface vegetation and the surface soil end member time series image for the target area to be identified for subsequent processing.

[0069] Step S102: Analyze the surface vegetation endmember time series images and the surface soil endmember time series images respectively to obtain the optimal fitting curves of the surface vegetation endmember time series and the surface soil endmember time series and the corresponding change characteristic parameters.

[0070] Specifically, adaptive fitting can be first performed on the surface vegetation endmember time series images and the surface soil endmember time series images to obtain the optimal fitting curves corresponding to the surface vegetation endmember time series and the surface soil endmember time series; and then the characteristic parameters of the optimal fitting curve are determined using the parameter selection rules matching the optimal fitting curve as the variation characteristic parameters.

[0071] The embodiment of the present invention forms a multi-type curve fitting optimization for the entire life cycle of abandoned farmland through adaptive fitting of time series, and provides characteristic parameters for abandoned farmland identification through analysis of different paths, so as to meet the reconstruction and modeling of the entire process of abandoned farmland under different climatic backgrounds (differences in natural recovery capacity of vegetation) and human interference (reclamation and fallow).

[0072] It is understandable that, considering the different evolution laws of vegetation and soil time series of abandoned farmland in different climatic zones, for example, due to the poor natural recovery of vegetation in arid areas, the vegetation and soil end-member time series of abandoned farmland are mostly presented as logistic curves; while in humid areas, due to the better natural vegetation recovery ability, the vegetation and soil end-member time series are mostly U-shaped; at the same time, considering that the reclamation behavior after abandonment of farmland can also be characterized by the U-shaped curve of vegetation and soil end-members. Therefore, when the embodiment of the present invention performs adaptive fitting on the time series image, the linear fitting, logical fitting and double logical fitting forms can be set to perform pixel-by-pixel fitting on the surface vegetation end-member time series image and the surface soil end-member time series image, respectively, to obtain a linear fitting curve, a logical fitting curve and a double logical fitting curve. Figure 2 As shown in FIG. 1 , the linear fitting curve, the logistic fitting curve and the double logistic fitting curve are schematically shown. Then, the residual root mean square error of each fitting curve is calculated, and the optimal fitting curve is determined based on the calculation results.

[0073] The residual root mean square error of the linear fitting curve is calculated as follows (1):

[0074]

[0075] The residual root mean square error of the logistic fitting curve is calculated as follows (2):

[0076]

[0077] The residual root mean square error of the dual logistic fitting curve is calculated as follows (3):

[0078]

[0079] Furthermore, after obtaining the residual root mean square error of each fitting curve, the fitting curve with the smallest residual root mean square error is selected as the optimal fitting curve, which is expressed as follows (4):

[0080] Model optimal =arg min(RMSE Linear ,RMSE Logistic ,RMSE DoubleLogistic ) (4)

[0081] Among them, β0 and β1 are the fitting parameters of the linear fitting curve; L, k, and t0 are the fitting parameters of the logistic fitting curve; L1, L2, k1, k2, t 0,1 and t 0,2 are the fitting parameters of the double logic fitting curve; i and x i are respectively the vegetation / soil end member time series value and year. In the embodiment of the present invention, the characteristic parameters of the optimal fitting curve are determined by using the parameter selection rule matching the optimal fitting curve as the change characteristic parameters, and the parameter selection rule can be specifically as follows: if the optimal fitting curve is a linear fitting curve, the linear trend of the linear fitting curve is used as the change characteristic parameter; if the optimal fitting curve is a logical fitting curve, the turning start time, turning end time and curve change amplitude in the logical fitting curve are used as the change characteristic parameters; if the optimal fitting curve is a dual logical fitting curve, the turning start time, the second turning time, the peak / valley arrival time, the peak / valley duration and the curve change amplitude in the dual logical fitting curve are used as the change characteristic parameters.

[0082] like Figure 3 As shown in FIG. 1 , they are schematic diagrams of the change characteristic parameters in the logic fitting curve and the dual logic fitting curve. In order to obtain these change characteristic parameters, the embodiment of the present invention determines them through the second-order derivative of the fitting curve, such as Figure 4 As shown in the figure, they are brief schematic diagrams of the first-order derivative and the second-order derivative of the fitting curve, respectively. The calculation formula of the first-order derivative is as follows (5):

[0083]

[0084] The formula for the second-order derivative is as follows (6):

[0085]

[0086] Where z = e at-b , a and b are the fitting parameters of the logic fitting curve respectively; the turning start time t0 of the logic fitting curve is the maximum value of the second-order derivative, and the turning end time t1 is the minimum value of the second-order derivative. Therefore, the formula for the curve change amplitude of the logic fitting curve and the dual logic fitting curve in the embodiment of the present invention is expressed as follows (7):

[0087] Mag=f(t1)-f(t0) (7)

[0088] The formula for the peak / valley duration of the dual logistic fitting curve is as follows (8):

[0089] Dur=t2-t1 (8)

[0090] Wherein, t0 is the turning start time; t1 is the turning end time or the peak / valley arrival time; t2 is the turning time again. It can be understood that the dual logic fitting curve is composed of two logic curves (the first logic fitting curve and the second logic fitting curve), and its parameter acquisition process is consistent with the process of the logic curve. Wherein, the turning start time t0 of the dual logic fitting curve is the maximum value of the second-order derivative of the first logic fitting curve, the peak / valley arrival time t1 is the minimum value of the second-order derivative of the first logic fitting curve, and the turning time t2 again is the maximum value of the second-order derivative of the second logic fitting curve.

[0091] Step S103, determining the type of abandoned farmland in the target area based on the optimal fitting curve of the target area and the corresponding change characteristic parameters.

[0092] It should be noted that the embodiment of the present invention not only uses the change of vegetation endmember time series, but also uses the change characteristics of soil endmember data series that interact with it to detect farmland abandonment. Therefore, the embodiment of the present invention mainly combines the change characteristic parameters of the two endmember time series to establish a rule-based farmland abandonment identification process, which is as follows:

[0093] If the optimal fitting curve of the target area is a logical fitting curve, the following formula (9) is used to determine the type of abandoned farmland in the target area:

[0094]

[0095] Among them, K is a positive empirical threshold; 1 represents the target area is a cultivated land abandonment area; 0 represents the target area is a non-cultivated land abandonment area; Mag v is the change range of the vegetation logic fitting curve in the target area; Mag s is the variation range of the soil logistic fitting curve in the target area.

[0096] It is understandable that the vegetation change range Mag v is a negative value, the soil change amplitude Mag s If it is a positive value, it means that vegetation degradation and soil exposure occurred in the process. The embodiment of the present invention determines the validity of this change feature by setting the change amplitude greater than a certain positive empirical threshold (K) as a standard. Therefore, based on the above characteristics, the abandoned cultivated land area can be determined.

[0097] If the optimal fitting curve of the target area is a dual logic fitting curve, the evolution characteristics of cultivated land are relatively complex, including three types of cultivated land abandonment, reclamation after abandonment, and fallow. Therefore, the embodiment of the present invention combines the land use and cover type data and the end member time series curve to establish rules and perform type classification. The specific formula is as follows (10):

[0098]

[0099] Among them, Mag v is the variation range of the vegetation double logic fitting curve in the target area; Mag s is the variation range of the soil double logistic fitting curve in the target area; Mag s >0 means that the soil double logistic fitting curve presents an inverted U-shape, otherwise Mag v <0 means that the vegetation double logic fitting curve presents a U-shape; 1 means that the target area is a potential cultivated land change area; 0 means that the target area is a non-cultivated land abandonment area. In other words, according to the preliminary judgment criteria of formula (10), the vegetation end member abundance is U-shaped and the soil end member abundance is inverted U-shaped, which represents the potential cultivated land change area.

[0100] Furthermore, when the target area is a potential cultivated land change area, the following formula (11) is used to determine the type of cultivated land abandonment in the target area:

[0101]

[0102] Among them, Dur v Duration of the valley value of the vegetation double logistic fitting curve in the target area; s is the peak duration of the soil dual logic fitting curve of the target area; T is the preset duration; 1 represents that the target area is a potential abandoned area; 0 represents that the target area is a fallow area. It can be understood that the judgment standard of this step is that the duration of the bottom of the vegetation end member abundance or the top of the soil end member abundance is greater than the preset duration, and the fallow area is eliminated. Preferably, the preset duration can be 2 years.

[0103] Furthermore, the embodiment of the present invention can also obtain the land use type of the target area in the later stage, and determine the type of farmland abandonment in the target area in the later stage based on the land use type, as shown in the following formula (12):

[0104]

[0105] Among them, LU later is the type of land use in the target area in the later period; Non_AG ​​is non-cultivated land; AG is cultivated land; 1 represents the target area as an abandoned area; 0 represents the target area as an abandoned and reclaimed area. The judgment standard of this step is to determine the situation of abandoned cultivated land and abandoned and reclaimed land according to the type of land use in the later period. If the land use and cover type in the later period is non-cultivated land, it is determined to be abandoned. If the land use and cover type in the later period is cultivated land, it is determined to be reclaimed after abandonment.

[0106] Based on the above steps, the multi-dimensional information full-process modeling of the farmland abandonment process is realized, and then the farmland abandonment identification with multi-dimensional information integration is carried out to accurately identify the target area as an abandoned area or a non-abandoned area, and can further identify the target area as a farmland abandonment area, a farmland reclamation area after abandonment, and a farmland fallow area.

[0107] An embodiment of the present invention provides a method for identifying abandoned farmland throughout its life cycle, including obtaining end-member time series images of surface vegetation and end-member time series images of surface soil in a target area; analyzing the end-member time series images of surface vegetation and end-member time series images of surface soil respectively to obtain the optimal fitting curves and corresponding change characteristic parameters of the end-member time series of surface vegetation and the end-member time series of surface soil; and determining the type of abandoned farmland in the target area based on the optimal fitting curves and corresponding change characteristic parameters of the target area. Through the present invention, a method for modeling and identifying abandoned farmland throughout its life cycle that integrates multi-dimensional end-member time series interactive information is provided. Through the detection of abandoned farmland based on the interactive mining of end-member time series information and the full-process modeling of multi-dimensional surface element information time series, it is helpful to refine the detection of the evolution characteristics of the entire process of abandoned farmland and improve the accuracy of abandoned farmland monitoring.

[0108] Further, as Figure 1 In a specific implementation, an embodiment of the present invention provides a device for identifying abandoned farmland throughout its life cycle, such as Figure 5 As shown, the device may include: a data acquisition module 510 , a data analysis module 520 and a type identification module 530 .

[0109] The data acquisition module 510 may be used to acquire the surface vegetation end-member time series images and the surface soil end-member time series images of the target area;

[0110] The data analysis module 520 can be used to analyze the surface vegetation end-member time series images and the surface soil end-member time series images respectively to obtain the optimal fitting curves and corresponding change characteristic parameters of the surface vegetation end-member time series and the surface soil end-member time series;

[0111] The type identification module 530 can be used to determine the type of abandoned farmland in the target area based on the optimal fitting curve of the target area and the corresponding change characteristic parameters.

[0112] Optionally, the data analysis module 520 may also be used to perform adaptive fitting on the surface vegetation endmember time series image and the surface soil endmember time series image to obtain the optimal fitting curve corresponding to the surface vegetation endmember time series and the surface soil endmember time series;

[0113] The characteristic parameters of the best fitting curve are determined by using the parameter selection rule matching the best fitting curve as the variation characteristic parameters.

[0114] Optionally, the data analysis module 520 can also be used to set the form of linear fitting, logical fitting and double logical fitting to fit the surface vegetation end member time series image and the surface soil end member time series image respectively, to obtain a linear fitting curve, a logical fitting curve and a double logical fitting curve;

[0115] The residual root mean square error of each fitting curve is calculated, and the best fitting curve is determined based on the calculation results.

[0116] Optionally, the data analysis module 520 may also be used to determine the best fitting curve using the following formula:

[0117]

[0118] Model optimal =argmin(RMSE Linear ,RMSE Logistic ,RMSE DoubleLogistic )

[0119] Among them, RMSE Linear is the residual root mean square error of the linear fitting curve; RMSE Logistic is the residual root mean square error of the logistic fitting curve; RMSE DoubleLogistic is the residual root mean square error of the double logistic fitting curve; Model optimal is the best fitting curve; β0 and β1 are the fitting parameters of the linear fitting curve; L, k, and t0 are the fitting parameters of the logistic fitting curve; L1, L2, k1, k2, t 0,1 and t 0,2 are the fitting parameters of the double logic fitting curve; i and x i are vegetation / soil endmember time series values ​​and years, respectively.

[0120] Optionally, the data analysis module 520 may also be used to use the linear trend of the linear fitting curve as a change characteristic parameter when the best fitting curve is a linear fitting curve;

[0121] When the best fitting curve is a logic fitting curve, the turning start time, turning end time and curve change amplitude in the logic fitting curve are used as change characteristic parameters;

[0122] When the best fitting curve is a double logic fitting curve, the turning start time, the second turning time, the peak / valley value arrival time, the peak / valley value duration and the curve change amplitude in the double logic fitting curve are used as change characteristic parameters.

[0123] Optionally, the change characteristic parameters of the logistic fitting curve and the dual logistic fitting curve are determined by second-order derivatives;

[0124] The formula for the second-order derivative is as follows:

[0125]

[0126] Where z = e at-b ; a and b are the fitting parameters of the logic fitting curve respectively; the turning start time t0 of the logic fitting curve is the maximum value of the second-order derivative, and the turning end time t1 is the minimum value of the second-order derivative; the turning start time t0 of the double logic fitting curve is the maximum value of the second-order derivative of the first logic fitting curve, the peak / valley value arrival time t1 is the minimum value of the second-order derivative of the first logic fitting curve, and the turning time t2 again is the maximum value of the second-order derivative of the second logic fitting curve;

[0127] and / or,

[0128] The formulas for the curve change amplitudes of the logical fitting curve and the double logical fitting curve are as follows:

[0129] Mag=f(t1)-f(t0)

[0130] and / or,

[0131] The formula for the peak / valley duration of the dual logistic fitting curve is as follows:

[0132] Dur=t2-t1

[0133] Among them, t0 is the starting time of the turning point; t1 is the ending time of the turning point or the peak / valley value arrival time; t2 is the next turning point time.

[0134] Optionally, the type identification module 530 may also be used to determine the type of abandoned farmland in the target area using the following formula when the optimal fitting curve of the target area is a logical fitting curve:

[0135]

[0136] Among them, K is a positive empirical threshold; 1 represents the target area is a cultivated land abandonment area; 0 represents the target area is a non-cultivated land abandonment area; Mag v is the change range of the vegetation logic fitting curve in the target area; Mag s The change range of the soil logistic fitting curve for the target area;

[0137] and / or,

[0138] When the optimal fitting curve of the target area is a double logic fitting curve, the following formula is used to determine the type of abandoned cultivated land in the target area:

[0139]

[0140] Among them, Mag v is the variation range of the vegetation double logic fitting curve in the target area; Mag s is the change range of the soil double logic fitting curve of the target area; 1 means that the target area is a potential cultivated land change area; 0 means that the target area is a non-cultivated land abandonment area;

[0141] When the target area is a potential cultivated land change area, the following formula is used to determine the type of cultivated land abandonment in the target area:

[0142]

[0143] Among them, Dur v Duration of the valley value of the vegetation double logistic fitting curve in the target area; s is the peak duration of the soil double logic fitting curve of the target area; T is the preset duration; 1 represents that the target area is a potential abandoned area; 0 represents that the target area is a fallow area.

[0144] Optionally, the type identification module 530 can also be used to obtain the land use type of the target area in the later stage, and determine the type of abandoned farmland in the target area in the later stage based on the land use type, and the formula is as follows:

[0145]

[0146] Among them, LU later is the later land use type of the target area; Non_AG ​​is non-cultivated land; AG is cultivated land; 1 represents that the target area is an abandoned area; 0 represents that the target area is a reclaimed area after abandonment.

[0147] It should be noted that for other corresponding descriptions of the functional modules involved in the device for identifying abandoned farmland throughout its life cycle provided by the embodiment of the present invention, reference can be made to Figure 1 The corresponding description of the method shown will not be repeated here.

[0148] Based on the above Figure 1 The method shown and Figure 5 The embodiment of the device shown in the figure, the embodiment of the present invention also provides a physical structure diagram of a computer device, such as Figure 6 As shown, the computer device may include a communication bus, a processor, a memory and a communication interface, and may also include an input / output interface and a display device, wherein each functional unit may communicate with each other through the bus. The memory stores a computer program, and the processor is used to execute the program stored in the memory and execute the steps of the method for identifying abandoned farmland throughout its life cycle described in the above embodiment.

[0149] Those skilled in the art can clearly understand that the specific working processes of the systems, devices, modules and units described above can refer to the corresponding processes in the aforementioned method embodiments, and for the sake of brevity, they are not further described here.

[0150] In addition, the functional units in various embodiments of the present invention may be physically independent of each other, or two or more functional units may be integrated together, or all functional units may be integrated into one processing unit. The above integrated functional units may be implemented in the form of hardware, or in the form of software or firmware.

[0151] Those skilled in the art can understand that if the integrated functional unit is implemented in the form of software and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention can essentially or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, which includes a number of instructions to enable a computing device (such as a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present invention when running the instructions. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), magnetic disk or optical disk and other media that can store program codes.

[0152] Alternatively, all or part of the steps of implementing the aforementioned method embodiments may be accomplished by hardware associated with program instructions (such as a computing device such as a personal computer, a server, or a network device), and the program instructions may be stored in a computer-readable storage medium. When the program instructions are executed by a processor of a computing device, the computing device executes all or part of the steps of the methods described in the embodiments of the present invention.

[0153] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that within the spirit and principles of the present invention, the technical solutions described in the aforementioned embodiments can still be modified, or some or all of the technical features therein can be replaced by equivalents. However, these modifications or replacements do not deviate from the protection scope of the present invention.

Claims

1. A method for identifying abandoned farmland throughout its life cycle, characterized in that: The method comprises: Obtain end-member time series images of surface vegetation and end-member time series images of surface soil in the target area; Analyzing the surface vegetation endmember time series image and the surface soil endmember time series image respectively to obtain the optimal fitting curves of the surface vegetation endmember time series and the surface soil endmember time series and the corresponding change characteristic parameters; Based on the optimal fitting curve of the target area and the corresponding change characteristic parameters, the type of abandoned farmland in the target area is determined.

2. The method according to claim 1, characterized in that: The analyzing the surface vegetation end-member time series image and the surface soil end-member time series image respectively to obtain the optimal fitting curves of the surface vegetation end-member time series and the surface soil end-member time series and the corresponding change characteristic parameters includes: Adaptively fitting the surface vegetation endmember time series image and the surface soil endmember time series image to obtain the optimal fitting curve corresponding to the surface vegetation endmember time series and the surface soil endmember time series; The characteristic parameters of the best fitting curve are determined by using a parameter selection rule that matches the best fitting curve as the change characteristic parameters.

3. The method according to claim 2, characterized in that The adaptive fitting is performed on the surface vegetation endmember time series image and the surface soil endmember time series image to obtain the optimal fitting curve corresponding to the surface vegetation endmember time series and the surface soil endmember time series, including: Setting the forms of linear fitting, logical fitting and double logical fitting to respectively fit the surface vegetation end-member time series image and the surface soil end-member time series image to obtain a linear fitting curve, a logical fitting curve and a double logical fitting curve; The residual root mean square error of each fitting curve is calculated, and the best fitting curve is determined based on the calculation results.

4. The method according to claim 3, characterized in that The step of calculating the residual root mean square error of each fitting curve and determining the best fitting curve based on the calculation result comprises: The best fit curve is determined using the following formula: Model optimal =argmin(RMSE Linear ,RMSE Logistic ,RMSE DoubleLogistic ) Among them, RMSE Linear is the residual root mean square error of the linear fitting curve; RMSE Logistic is the residual root mean square error of the logistic fitting curve; RMSE DoubleLogistic is the residual root mean square error of the double logistic fitting curve; Model optimal is the best fitting curve; β0 and β1 are the fitting parameters of the linear fitting curve; L, k, and t0 are the fitting parameters of the logistic fitting curve; L1, L2, k1, k2, t 0,1 and t 0,2 are the fitting parameters of the double logic fitting curve; i and x i are vegetation / soil endmember time series values ​​and years, respectively.

5. The method according to claim 3, characterized in that: The method of determining the characteristic parameters of the best fitting curve by using the parameter selection rule matching the best fitting curve as the change characteristic parameters includes: If the optimal fitting curve is a linear fitting curve, the linear trend of the linear fitting curve is used as the change characteristic parameter; If the optimal fitting curve is a logic fitting curve, the turning start time, turning end time and curve change amplitude in the logic fitting curve are used as the change characteristic parameters; If the optimal fitting curve is a dual logic fitting curve, the turning start time, the second turning time, the peak / valley arrival time, the peak / valley duration time and the curve change amplitude in the dual logic fitting curve are used as the change characteristic parameters.

6. The method according to claim 1, characterized in that The change characteristic parameters of the logic fitting curve and the dual logic fitting curve are determined by second-order derivatives; The formula of the second-order derivative is expressed as follows: Where z = e at-b ; a and b are fitting parameters of the logic fitting curve respectively; the turning start time t0 of the logic fitting curve is the maximum value of the second-order derivative, and the turning end time t1 is the minimum value of the second-order derivative; the turning start time t0 of the double logic fitting curve is the maximum value of the second-order derivative of the first logic fitting curve, the peak / valley value arrival time t1 is the minimum value of the second-order derivative of the first logic fitting curve, and the turning time t2 again is the maximum value of the second-order derivative of the second logic fitting curve; and / or, The formulas for the curve variation ranges of the logic fitting curve and the dual logic fitting curve are expressed as follows: Mag=f(t1)-f(t0) and / or, The formula for the peak / valley duration of the dual logic fitting curve is expressed as follows: Dur=t2-t1 Among them, t0 is the starting time of the turning point; t1 is the ending time of the turning point or the peak / valley value arrival time; t2 is the next turning point time.

7. The method according to claim 5, characterized in that The step of determining the type of abandoned farmland in the target area based on the optimal fitting curve and the corresponding change characteristic parameters of the target area includes: If the optimal fitting curve of the target area is a logical fitting curve, the type of abandoned farmland in the target area is determined using the following formula: Among them, K is a positive empirical threshold; 1 represents that the target area is an abandoned farmland area; 0 represents that the target area is an abandoned non-farmland area; Mag v is the change range of the vegetation logic fitting curve in the target area; Mag s is the change range of the soil logic fitting curve of the target area; and / or, If the optimal fitting curve of the target area is a dual logic fitting curve, the type of abandoned farmland in the target area is determined using the following formula: Among them, Mag v is the variation range of the vegetation double logic fitting curve in the target area; Mag s is the change amplitude of the soil double logic fitting curve of the target area; 1 represents that the target area is a potential cultivated land change area; 0 represents that the target area is a non-cultivated land abandonment area; When the target area is a potential cultivated land change area, the type of abandoned cultivated land in the target area is determined using the following formula: Among them, Dur v Dur is the valley duration of the vegetation double logic fitting curve in the target area; s is the peak duration of the soil double logic fitting curve of the target area; T is the preset duration; 1 represents that the target area is a potential abandoned area; 0 represents that the target area is a fallow area.

8. The method according to claim 7, characterized in that The method further comprises: The land use type of the target area in the later stage is obtained, and the type of abandoned farmland in the target area in the later stage is determined based on the land use type. The formula is as follows: Among them, LU later is the later land use type of the target area; Non_AG ​​is non-cultivated land; AG is cultivated land; 1 represents that the target area is an abandoned area; 0 represents that the target area is a reclaimed area after abandonment.

9. A device for identifying abandoned farmland throughout its life cycle, characterized in that: The device comprises: A data acquisition module is used to acquire end-member time series images of surface vegetation and end-member time series images of surface soil in a target area; A data analysis module, used to analyze the surface vegetation end-member time series image and the surface soil end-member time series image respectively, to obtain the optimal fitting curves of the surface vegetation end-member time series and the surface soil end-member time series and the corresponding change characteristic parameters; The type identification module is used to determine the type of abandoned farmland in the target area based on the optimal fitting curve of the target area and the corresponding change characteristic parameters.

10. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the method for identifying abandoned farmland throughout its life cycle as described in any one of claims 1 to 8 is implemented.

Images