An adaptive feature soil salinity estimation method and device

Abstract

The application discloses a kind of self-adapting feature optimization's soil salt content estimation method and device.Therein, method includes: using quality evaluation data to remove cloud and water body in real-time acquisition Landsat-9 image band data, obtain the pre-processed Landsat-9 image band data;From the pre-processed Landsat-9 image band data, extract the real-time feature element set corresponding to the pre-set feature element set;Real-time feature element set is respectively input into the random forest soil salt content estimation model and LightGBM soil salt content estimation model that are constructed in advance, output first soil salt content estimation result and second soil salt content estimation result;First soil salt content estimation result and second soil salt content estimation result are input into the soil salt content calculation formula that contains temperature adjustment factor and are constructed in advance, obtain final soil salt content estimation result.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of soil salinity estimation, and more particularly, to a soil salinity estimation method and device with adaptive features. BACKGROUND

[0002] Traditional soil salinization extraction methods require field investigation and laboratory testing to obtain relevant information on salinization. In contrast, satellite remote sensing observations can monitor soil salinity in a large area at a relatively low cost. With the development of computer technology, the construction of soil salinization remote sensing estimation models has gradually shifted from traditional statistical methods to machine learning algorithms. Although machine learning algorithms have obvious advantages in non-linear modeling and automatic feature extraction, small sample size machine learning soil salinization remote sensing monitoring often cannot fully represent the complexity of the entire region. The model may only remember the specific noise of limited data during training, resulting in insufficient generalization ability on unknown data and overfitting problems. Small sample data are more difficult to ensure the robustness of the model when facing noise, further reducing the accuracy of the model and the accuracy of soil salinity estimation. SUMMARY

[0003] To overcome the deficiencies of the prior art, the present application provides a soil salinity estimation method and device with adaptive features.

[0004] According to one aspect of the present application, a soil salinity estimation method with adaptive features is provided, comprising:

[0005] The quality evaluation data is used to remove clouds and water bodies in the real-time acquired Landsat-9 image band data to obtain pre-processed Landsat-9 image band data;

[0006] The pre-processed Landsat-9 image band data is used to extract a real-time feature element set corresponding to a pre-set feature element set;

[0007] The real-time feature element set is input into a pre-constructed random forest soil salinity estimation model and a LightGBM soil salinity estimation model, respectively, to output a first soil salinity estimation result and a second soil salinity estimation result;

[0008] The first soil salinity estimation result, the second soil salinity estimation result, the site temperature data, and a pre-constructed soil salinity calculation formula containing a temperature adjustment factor are used to calculate a final soil salinity estimation result.

[0009] According to another aspect of the present application, a soil salinity estimation device with adaptive features is provided, comprising:

[0010] The removing module is used for removing clouds and water bodies in the real-time acquired Landsat-9 image band data by using the quality evaluation data, so as to obtain the preprocessed Landsat-9 image band data.

[0011] The extracting module is used for extracting a real-time feature element set corresponding to a pre-set feature element set from the preprocessed Landsat-9 image band data.

[0012] The estimating module is used for inputting the real-time feature element set into a pre-constructed random forest soil salt estimation model and a LightGBM soil salt estimation model respectively, and outputting a first soil salt estimation result and a second soil salt estimation result.

[0013] The calculating module is used for calculating a final soil salt estimation result according to the first soil salt estimation result, the second soil salt estimation result, site temperature data and a pre-constructed soil salt calculation formula containing a temperature adjustment factor.

[0014] According to another aspect of the present application, a computer readable storage medium is provided, which stores a computer program for executing the method according to any one of the above aspects of the present application.

[0015] According to another aspect of the present application, an electronic device is provided, which comprises a processor, a memory for storing executable instructions of the processor, and the processor is configured to read the executable instructions from the memory and execute the instructions to implement the method according to any one of the above aspects of the present application.

[0016] Therefore, the present application provides a weather factor adjusted adaptive feature soil salt estimation method: by alternately using the feature importance sorting of random forest (RF) and LightGBM (LGBM), combining the algorithm advantages of the two to perform double model importance cross-validation and fusion, dynamically filtering feature subsets and synchronously optimizing model hyperparameters, and performing soil salt estimation, the estimation accuracy of soil salt is greatly improved. BRIEF DESCRIPTION OF DRAWINGS

[0017] The exemplary embodiments of the present application can be more completely understood by reference to the following drawings:

[0018] Figure 1 is a flowchart of the adaptive feature soil salt estimation method provided by an exemplary embodiment of the present application;

[0019] Figure 2 is a structural schematic diagram of the adaptive feature soil salt estimation device provided by an exemplary embodiment of the present application;

[0020] Figure 3is a structure of an electronic device provided by an exemplary embodiment of the present application. DETAILED DESCRIPTION

[0021] Hereinafter, exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. It should be apparent that these described embodiments are merely exemplary of the present application and should not be considered limiting the scope of the present application. Therefore, the disclosure of these exemplary embodiments is intended to be illustrative, but not limiting, of the scope of the present application.

[0022] It should be noted that the relative arrangement of the components and steps, the numerical expressions, and numerical values set forth in these embodiments are not limiting of the scope of the application unless otherwise specifically stated.

[0023] Those skilled in the art can understand that the terms "first", "second" and the like in the embodiments of the present application are only used to distinguish different steps, devices or modules, and do not represent any specific technical meaning, nor indicate their logical order.

[0024] It should also be understood that in the embodiments of the present application, "a plurality of" can mean two or more, and "at least one" can mean one, two or more.

[0025] It should also be understood that for any component, data or structure mentioned in the embodiments of the present application, unless specifically limited or unless the context clearly indicates otherwise, it can be understood as one or more.

[0026] In addition, the term "and / or" in the present application is only a description of the association relationship between the associated objects, which means that there can be three relationships, for example, A and / or B can represent the existence of A alone, the existence of A and B together, and the existence of B alone. In addition, the character " / " in the present application generally represents an "or" relationship between the associated objects.

[0027] It should also be understood that the description of each embodiment of the present application emphasizes the differences between each embodiment, and the same or similar parts can be referred to each other, and for the sake of brevity, will not be repeated.

[0028] At the same time, it should be understood that in order to facilitate description, the size of each part shown in the drawings is not drawn according to the actual proportion relationship.

[0029] The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way limiting of the application or its use or functionality.

[0030] Techniques, methods, and devices known to those of ordinary skill in the relevant art can not be discussed in detail herein, but should be considered as part of the specification for purposes of patentability.

[0031] It should be noted that like reference numerals and characters refer to like items throughout the attached drawings and alternative embodiments thereof. Commonly, once an item is defined in one drawing, it need not be discussed further in subsequent drawings.

[0032] Embodiments of the present application can be applied to terminal devices, computer systems, servers, and other electronic devices, which can operate with many other general-purpose or special-purpose computing system environments or configurations. Examples of well-known terminal devices, computing systems, environments, and / or configurations suitable for use with terminal devices, computer systems, servers, and other electronic devices include, but are not limited to, personal computer systems, server computer systems, thin clients, thick clients, handheld or laptop devices, microprocessor-based systems, set-top boxes, programmable consumer electronics, network personal computers, minicomputer systems, mainframe computer systems, and distributed cloud computing technology environments comprising any of the above systems, and the like.

[0033] Terminal devices, computer systems, servers, and other electronic devices can be described in the general context of computer system-executable instructions, such as program modules, being executed by a computer system. Generally, program modules can include routines, programs, objects, components, logic, data structures, and the like, which perform particular tasks or implement particular abstract data types. Computer systems / servers can be implemented in a distributed cloud computing environment, in which tasks are performed by remote processing devices that are linked through a communications network. In a distributed cloud computing environment, program modules can be located in local or remote computer system storage media including storage devices.

[0034] Exemplary method

[0035] Figure 1 is a flowchart of an adaptive feature soil salinity estimation method provided by an exemplary embodiment of the present application. The present embodiment can be applied to electronic devices, such as Figure 1 As shown in the figure, the adaptive feature soil salinity estimation method 100 includes the following steps:

[0036] Step 101, using quality evaluation data to remove clouds and water bodies in the real-time acquired Landsat-9 image band data, to obtain pre-processed Landsat-9 image band data;

[0037] Step 102, extracting a real-time feature element set corresponding to a pre-set feature element set from the pre-processed Landsat-9 image band data;

[0038] Step 103, inputting the real-time feature element set into a pre-constructed random forest soil salinity estimation model and a LightGBM soil salinity estimation model respectively, and outputting a first soil salinity estimation result and a second soil salinity estimation result;

[0039] Step 104, according to the first soil salt estimation result, the second soil salt estimation result, the site temperature data and the pre-constructed soil salt calculation formula containing the temperature adjustment factor, the final soil salt estimation result is calculated.

[0040] Specifically, in view of the technical problems existing in the background art, the present application proposes a soil salt estimation method based on adaptive optimal feature selection (AOFS): by alternately using the feature importance ranking of random forest (RF) and LightGBM (LGBM), combining the algorithm advantages of the two to perform double model importance cross-validation and fusion, dynamically selecting feature subsets and synchronously optimizing model hyperparameters, and performing soil salt estimation. The specific implementation is as follows:

[0041] 1. Landsat-9 image band data and quality assessment data (Quality Assessment, QA), using QA data to remove clouds and water bodies, extracting and calculating 37 feature element sets , combined with field investigation and laboratory detection of measured data (including latitude and longitude information and conductivity data), using measured latitude and longitude and time information to match Landsat-9 data for data labeling, and matching spatial and temporal adjacent weather station temperature data (used to calculate temperature adjustment factor) to obtain a matched sample set, wherein the conductivity data is used as the true value, and the sample set is divided into a training set and a validation set in a ratio of 7:3.

[0042] 、

[0043] 、

[0044]

[0045] In the formula, 、 、 、 、 、 are the reflectances of red, green, blue, near-infrared, short-wave infrared 1, and short-wave infrared 2, respectively.

[0046] 2. Use the feature importance ranking of random forest (RF) and LightGBM (LGBM) to adaptively adjust the feature selection strategy according to the model performance, the initial feature element set f = F , the models of RF and LGBM and , the initial root mean square error of the validation set corresponding to the model , :

[0047] (1) Using the training set, use network search to optimize RF and LGBM hyperparameters to obtain the model and , and use the validation set to obtain and , if , then , , , , jump to step (2); if , the model , , , unchanged, jump to step (2);

[0048] (2) If the number of characteristic element set is not greater than 5, jump to (4), if the number of characteristic element set is greater than 5, jump to (3);

[0049] (3) Get the characteristic element set The corresponding model , feature importance ranking, the importance of RF and LGBM and cross fusion to get fusion importance , according to the fusion importance, remove the last feature, that is , update , return to (1).

[0050] Dynamic weight fusion of element importance:

[0051] Among them, is the model weight factor, which is determined by the performance of RF and LGBM model, and the model with better performance is given higher weight. According to the validation set RMSE , .

[0052] (4) Output the optimal model , , , and the corresponding final iteration characteristic element set , output the final weight factor :

[0053]

[0054] 3、Characteristic element , Landsat-9 all input data, using model 、 Estimate soil salinity and , combined , calculate the simulated soil salinity:

[0055]

[0056] 4、According to the simulated soil salinity , add temperature adjustment factor , the least square fitting of linear relationship a and b :

[0057] Wherein, Temperature adjustment factor; T For matching space-time adjacent weather station temperature data, For all matching point temperature mean. SS For the true result. a and b Fitting parameters in linear relationship.

[0058] 5、Landsat-9 image band data and quality assessment data (Quality Assessment, QA), using QA data to cloud and water, get 、 、 、 、 、 Calculate the final two-dimensional feature elements in the feature element set, according to 、 Estimate the two-dimensional simulated soil salinity, and then calculate the final two-dimensional soil salinity result according to the linear relationship in step 4.

[0059] Final calculation formula:

[0060]

[0061] Therefore, the present application proposes a kind of meteorological factor adjusted self-adapting feature soil salinity estimation method: by using the feature importance ranking of random forest (RF) and LightGBM (LGBM) alternately, cross validation and fusion of double model importance are carried out combining the algorithm advantages of two, dynamically filter feature subset and simultaneously optimize model hyperparameter, carry out soil salinity estimation. Greatly improve the estimation accuracy of soil salinity.

[0062] Exemplary apparatus

[0063] Figure 2 is a structural schematic diagram of an adaptive feature soil salinity estimation device provided by an exemplary embodiment of the present application. As shown in Figure 2 , the device 200 includes:

[0064] The removal module 210 is configured to remove clouds and water bodies in the real-time acquired Landsat-9 image band data by using the quality assessment data, to obtain pre-processed Landsat-9 image band data.

[0065] The extraction module 220 is configured to extract a real-time feature element set corresponding to a pre-set feature element set from the pre-processed Landsat-9 image band data.

[0066] The estimation module 230 is configured to input the real-time feature element set into a pre-constructed random forest soil salinity estimation model and a LightGBM soil salinity estimation model, respectively, to output a first soil salinity estimation result and a second soil salinity estimation result.

[0067] The calculation module 240 is configured to calculate a final soil salinity estimation result according to the first soil salinity estimation result, the second soil salinity estimation result, site temperature data, and a pre-constructed soil salinity calculation formula containing a temperature adjustment factor.

[0068] Exemplary electronic device

[0069] Figure 3 is a structure of an electronic device provided by an exemplary embodiment of the present application. As shown in Figure 3 , the electronic device 30 includes one or more processors 31 and a memory 32.

[0070] The processor 31 can be a central processing unit (CPU) or other forms of processing units with data processing capability and / or instruction execution capability, and can control other components in the electronic device to perform desired functions.

[0071] The memory 32 can include one or more computer program products that can include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory, for example, can include random access memory (RAM), and / or a cache, etc. The non-volatile memory, for example, can include read-only memory (ROM), hard disk, flash memory, etc. One or more computer program instructions can be stored on the computer-readable storage media, and the processor 31 can execute the program instructions to implement the methods of the software programs of the various embodiments of the present application described above and / or other desired functions. In one example, the electronic device can further include an input device 33 and an output device 34, which are interconnected through a bus system and / or other forms of connection mechanisms (not shown).

[0072] In addition, the input device 33 can include, for example, a keyboard, a mouse, etc.

[0073] The output device 34 can output various information to the outside. The output device 34 can include, for example, a display, a speaker, a printer, a communication network and a remote output device connected thereto, etc.

[0074] Of course, in order to simplify, Figure 3 Only some of the components of the electronic device related to the present application are shown in FIG. 1, and components such as a bus, an input / output interface, etc. are omitted. In addition, the electronic device can include any other appropriate components according to the specific application.

[0075] Exemplary computer program product and computer readable storage medium

[0076] In addition to the methods and devices described above, embodiments of the present application can also be a computer program product including computer program instructions that, when executed by a processor, cause the processor to perform the steps of the methods according to various embodiments of the present application described in the "Exemplary Methods" section of the present specification.

[0077] The computer program product can be written in any combination of one or more programming languages, including an object-oriented programming language such as Java, C++, etc., and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code can be executed entirely on the user computing device, partially on the user device, as a standalone software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0078] In addition, an embodiment of the present application can also be a computer readable storage medium, having stored thereon computer program instructions which, when executed by a processor, cause the processor to carry out the steps described in the above "Exemplary Method" section of this specification of the methods according to various embodiments of the present application.

[0079] The computer readable storage medium can be any combination of one or more computer readable medium(s). The computer readable medium can be a computer readable signal medium or a computer readable storage medium. A computer readable storage medium can be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, or apparatus or device, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of the computer readable storage medium include an electrical connection having one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0080] The above describes the basic principles of the present application in conjunction with specific embodiments, but it should be noted that the advantages, benefits, effects and the like mentioned in the present application are only examples and are not limiting, and these advantages, benefits, effects and the like cannot be considered as necessary for each embodiment of the present application. In addition, the above specific details are only for the purpose of example and for the purpose of understanding, and are not limiting, and the above details do not limit the present application to the above specific details.

[0081] Each embodiment in the specification is described in a progressive manner, and each embodiment focuses on the differences from other embodiments, and the same or similar parts between each embodiment can be mutually referred to. For system embodiments, since they basically correspond to method embodiments, the description is relatively simple, and the relevant parts can be referred to the part of the method embodiment.

[0082] The block diagrams of the devices, systems, apparatuses, systems involved in the present application are only illustrative examples and are not intended to require or imply that the connections, arrangements, configurations must be as shown in the block diagrams. As those skilled in the art will recognize, these devices, systems, apparatuses, systems can be connected, arranged, configured in any manner. Words such as "include", "contain", "have" and the like are open-ended words, mean "including but not limited to", and can be used interchangeably. The words "or" and "and" used herein mean the word "and / or", and can be used interchangeably unless the context clearly indicates otherwise. The word "such as" used herein means the phrase "such as but not limited to", and can be used interchangeably.

[0083] The methods and systems of the present application can be implemented in a number of ways. For example, the methods and systems of the present application can be implemented via software, hardware, firmware, or any combination of software, hardware, and firmware. The above described order of steps for the methods is merely illustrative, and the steps of the methods of the present application are not limited to the order specifically described above unless otherwise specifically stated. Furthermore, in some embodiments, the present application can also be implemented as a program recorded on a recording medium, which includes machine readable instructions for implementing the methods according to the present application. Thus, the present application also covers recording media storing programs for executing the methods according to the present application.

[0084] It is also important to note that the systems, devices and methods of the present application can be embodied in a variety of forms without departing from the spirit of the application. Specifically, the systems, devices and methods of the present application can be implemented using hardware, software, firmware, or any combination thereof. In some embodiments, the systems, devices and methods of the present application can be implemented as a program tangibly embodied on a program carrier. It is therefore intended that the present application covers all such variations and modifications that fall within the scope of the application. It is also intended that the preambles of the claims be interpreted even broader to encompass all such variations and modifications.

[0085] The above description has been presented for the purposes of illustration and description. Further, this description is not intended to limit the embodiments of the present application to the forms disclosed herein. Although several example aspects and embodiments have been discussed, it should be recognized that various modifications, substitutions, changes, additions and rearrangements can be made by those skilled in the art without departing from the spirit of the application.

Claims

1. A method for estimating soil salinity using adaptive features, characterized in that, include: Clouds and water bodies were removed from the real-time acquired Landsat-9 image band data using quality assessment data to obtain preprocessed Landsat-9 image band data. Extract the real-time feature set corresponding to the pre-defined feature set from the pre-processed Landsat-9 image band data; The real-time feature set is input into the pre-constructed random forest soil salinity estimation model and the LightGBM soil salinity estimation model, respectively, and the first soil salinity estimation result and the second soil salinity estimation result are output. The final soil salinity estimation result is calculated based on the first soil salinity estimation result, the second soil salinity estimation result, the station temperature data, and the pre-constructed soil salinity calculation formula containing temperature adjustment factors. The construction process of the feature set, the random forest soil salinity estimation model, and the LightGBM soil salinity estimation model is as follows: Historical Landsat-9 image band data were de-clouded and water body removed using historical quality assessment data to obtain pre-processed historical Landsat-9 image band data. An initial feature set F and the true conductivity output are extracted from the preprocessed historical Landsat-9 image band data as a matching sample set. An adaptive feature selection strategy is adopted. Based on the matched sample set, the feature importance of the random forest model and the LightGBM model is ranked and iteratively trained to obtain the final feature set, the random forest soil salinity estimation model, the LightGBM soil salinity estimation model, and the model weight factors, including: Step 1: Divide the matching sample set into a training set and a validation set; Step 2: Initialize the Random Forest Model and LightGBM model and initialize the feature set. f=F The initial root mean square error of the corresponding model's validation set. , ; Step 3: Using the training set, simultaneously optimize the random forest model using network search. RF and LightGBM model LGBM Hyperparameters are used to obtain the model and The root mean square error of the two models was obtained using the validation set. and ,like Then let , , , Proceed to step 4; if Then the model , , , No change, proceed to step 4; Step 4: If the feature set If the number is no more than 5, proceed to step 6. If the feature set If the number is greater than 5, proceed to step 5; Step 5: Importance of Random Forest (RF) and LightGBM (LGBM) models and Cross-integration reveals the importance of integration. The last feature is removed after sorting by fusion importance to obtain the updated feature set. ,renew Return to step 1; Step 6: Output the optimal random forest soil salinity estimation model LightGBM Soil Salinity Estimation Model Root mean square error of the final random forest The final root mean square error of LightGBM The feature set of the final iteration and the final model weight factors ; Key feature elements corresponding to the matched sample set are input into the random forest soil salinity estimation model. LightGBM Soil Salinity Estimation Model Estimated soil salinity of random forest samples And the soil salinity estimation results of the LightGBM sample ; according to and Calculation of simulated soil salinity ; Temperature regulation factor added to the simulated soil salinity A linear simulation was performed to obtain a formula for calculating soil salinity. The calculation expression is: The temperature regulation factor The expression is: The formula for calculating soil salinity is: In the formula, SS The actual results were used in the linear simulation for the final estimated soil salinity. T To match temperature data from nearby meteorological stations, The average temperature of all matched points. a and b These are the fitting parameters in the linear relationship.

2. The method according to claim 1, characterized in that, The initial feature set includes: 、 、 In the formula, , , , , , The reflectances are red, green, blue, near-infrared, short-wave infrared 1, and short-wave infrared 2, respectively.

3. The method according to claim 1, characterized in that, The expression for calculating the importance of fusion is: in, The model weight factor is calculated as follows: In the formula, For the first The feature set of the next iteration.

4. A soil salinity estimation device for implementing the adaptive features of the method according to any one of claims 1-3, characterized in that, include: The removal module is used to remove clouds and water bodies from the real-time acquired Landsat-9 image band data using quality assessment data, so as to obtain preprocessed Landsat-9 image band data. The extraction module is used to extract the real-time feature set corresponding to the pre-defined feature set from the pre-processed Landsat-9 image band data. The estimation module is used to input the real-time feature set into the pre-constructed random forest soil salinity estimation model and the LightGBM soil salinity estimation model, respectively, and output the first soil salinity estimation result and the second soil salinity estimation result. The calculation module is used to calculate the final soil salinity estimation result based on the first soil salinity estimation result, the second soil salinity estimation result, the station temperature data, and a pre-constructed soil salinity calculation formula that includes a temperature adjustment factor.

5. A computer-readable storage medium, characterized in that, The storage medium stores a computer program for performing the method described in any one of claims 1-3.

6. An electronic device, characterized in that, The electronic device includes: processor; Memory used to store the processor's executable instructions; The processor is configured to read the executable instructions from the memory and execute the instructions to implement the method described in any one of claims 1-3.

Images