Irrigation area multi-type crop identification method and system based on active and passive time sequence remote sensing cooperation

Abstract

The invention provides an irrigation area multi-type crop identification method and system based on active and passive time sequence remote sensing cooperation. The method comprises the steps of obtaining active remote sensing data and passive remote sensing data; preprocessing the active remote sensing data to obtain an SAR remote sensing image; preprocessing the passive remote sensing data to obtain an optical remote sensing image; calculating a normalized vegetation index based on the optical remote sensing image; superposing the SAR remote sensing image, the optical remote sensing image and the normalized vegetation index according to a time sequence, and constructing a remote sensing data time sequence feature set; constructing an irrigation area multi-type crop sample point set; based on the remote sensing data time sequence feature set and the irrigated area multi-type crop sample point set, training a plurality of base classifiers, and then utilizing the plurality of base classifiers to identify the irrigated area multi-type crops; and carrying out integrated classification on the identification results of the plurality of base classifiers to obtain a classification result of the multi-type crops in the irrigation area.

Description

Technical Field

[0001] The present application relates to the technical field of crop identification, and in particular to a method and system for identifying multiple types of crops in irrigation areas using active and passive time-series remote sensing collaboration. Background Art

[0003] Irrigation areas usually have large geographical spans and widely distributed cultivated land. Differences in climate change and field management can easily lead to different time-varying information of the same crop in different plots, and there are generally fragmented plots and diverse crop types. Therefore, it is challenging to quickly and accurately obtain agricultural information of crops in irrigation areas. Conventional methods mainly rely on manual surveys and statistical reports, which have shortcomings such as high cost and low efficiency when applied on a large scale, and it is difficult to effectively adapt to the emergency decision-making needs of management departments. Remote sensing technology has the characteristics of wide data coverage and strong timeliness, and has become the main technical means for obtaining crop agricultural information. Facing irrigation area scenarios, when using remote sensing technology to obtain large-scale crop agricultural information, it is mainly faced with interference from factors such as cloudy and rainy weather, "different objects with the same spectrum" and "same object with different spectrum", resulting in low accuracy of agricultural information and weak classification generalization. Summary of the invention

[0004] The embodiments of the present application provide a method and system for identifying multiple types of crops in irrigation areas by coordinating active and passive time-series remote sensing to accurately obtain agricultural information of crops in irrigation areas.

[0005] In a first aspect, an embodiment of the present application provides a method for identifying multiple types of crops in an irrigation area using active and passive time-series remote sensing collaboration, including:

[0006] Acquire active remote sensing data and passive remote sensing data;

[0007] Preprocessing the active remote sensing data to obtain SAR remote sensing images;

[0008] Preprocessing the passive remote sensing data to obtain an optical remote sensing image;

[0009] Calculate the Normalized Difference Vegetation Index based on optical remote sensing images;

[0010] The SAR remote sensing images, optical remote sensing images and normalized vegetation index are superimposed in chronological order to construct a time series feature set of remote sensing data;

[0011] Construct a sample point set of multiple types of crops in irrigation areas;

[0012] Based on the remote sensing data time series feature set and the sample point set of multiple types of crops in the irrigation area, multiple base classifiers are trained, and then the multiple types of crops in the irrigation area are respectively identified using the multiple base classifiers;

[0013] The recognition results of multiple base classifiers are integrated and classified to obtain the classification results of multiple types of crops in irrigation areas.

[0014] In a feasible implementation, the step of acquiring active remote sensing data and passive remote sensing data includes:

[0015] Active remote sensing data and passive remote sensing data are screened out according to the vector range of the irrigation area.

[0016] In a feasible implementation, the preprocessing of the active remote sensing data to obtain a SAR remote sensing image includes:

[0017] The active remote sensing data was subjected to terrain correction, Refined Lee filtering, splicing, cropping and standard deviation normalization to form SAR remote sensing images covering the entire irrigation area from May to October;

[0018] The SAR remote sensing image is a multi-view image of ground distance in an interferometric wide-band mode, the polarization modes are VV polarization and VH polarization, and the image spatial resolution is 10m.

[0019] In a feasible implementation, the preprocessing of the passive remote sensing data to obtain an optical remote sensing image includes:

[0020] The passive remote sensing data were declouded using the QA60 band, and then optical remote sensing images covering the entire irrigation area from May to October were constructed through splicing, monthly median synthesis, cropping, and standard deviation normalization.

[0021] The optical remote sensing image is an L2A surface reflectance product, and the image used is composed of 10 spectral bands.

[0022] In a feasible implementation, the calculating of the normalized vegetation index based on the optical remote sensing image includes:

[0023] The formula for calculating the normalized vegetation index is:

[0024]

[0025] Where NDVI represents the normalized difference vegetation index, ρ R represents the red band reflectance of the image, ρ NIR Represents the image near-infrared band reflectance.

[0026] In a feasible implementation, the construction of a sample point set of multiple types of crops in an irrigation area includes:

[0027] Crop samples are obtained through a combination of irrigation area surveys, high-resolution image identification, and irrigation area data.

[0028] In a feasible implementation, the method of training multiple base classifiers based on the remote sensing data time series feature set and the irrigation area multi-type crop sample point set, and then using the multiple base classifiers to respectively identify the irrigation area multi-type crops includes:

[0029] The remote sensing data time series feature set is respectively input into a random forest base classifier, a rotating forest base classifier and a particle swarm optimized support vector machine base classifier, and the crop sample points in the irrigation area multi-type crop sample point set are divided into a training set and a test set in a 1:1 ratio, and the random forest base classifier, the rotating forest base classifier and the particle swarm optimized support vector machine base classifier are respectively trained and tested using the training set and the test set.

[0030] In a feasible implementation, the identification results of multiple base classifiers are integrated and classified to obtain the classification results of multiple types of crops in the irrigation area, including:

[0031] Based on the recognition results of multiple base classifiers, the overall classification accuracy of each base classifier is calculated respectively;

[0032] The recognition results of multiple base classifiers are voted by majority, and the one with the highest number of votes is determined as the classification result of multiple types of crops in the irrigation area. If the votes are the same, the recognition result of the base classifier with the highest overall classification accuracy is determined as the classification result of multiple types of crops in the irrigation area.

[0033] The calculation formula for the overall classification accuracy is:

[0034]

[0035] In the formula, OA represents the overall classification accuracy, M represents the confusion matrix, n represents the number of categories, N represents the total number of test samples, and M uu Represents the number of samples of category u that are correctly identified as u, u∈[1,n].

[0036] In a feasible implementation, the active and passive time-series remote sensing coordinated irrigation area multi-type crop identification method further includes:

[0037] An accuracy evaluation is performed on the classification results.

[0038] In a second aspect, an embodiment of the present application provides an irrigation area multi-type crop identification system using active and passive time-series remote sensing collaboration, which uses the irrigation area multi-type crop identification method using active and passive time-series remote sensing collaboration as described in the first aspect, including:

[0039] A remote sensing data acquisition module is used to acquire active remote sensing data and passive remote sensing data;

[0040] A preprocessing module, used for preprocessing the active remote sensing data to obtain SAR remote sensing images, and for preprocessing the passive remote sensing data to obtain optical remote sensing images;

[0041] Normalized difference vegetation index calculation module, used to calculate the normalized difference vegetation index based on optical remote sensing images;

[0042] The remote sensing data temporal feature set construction module is used to superimpose SAR remote sensing images, optical remote sensing images and normalized vegetation index in chronological order to construct the remote sensing data temporal feature set;

[0043] Sample point set construction module, used to construct sample point sets of multiple types of crops in irrigation areas;

[0044] A crop identification module is used to train multiple base classifiers based on a remote sensing data time series feature set and a sample point set of multiple types of crops in the irrigation area, and then use the multiple base classifiers to respectively identify the multiple types of crops in the irrigation area;

[0045] The integrated classification module is used to integrate the recognition results of multiple base classifiers to obtain the classification results of multiple types of crops in the irrigation area.

[0046] The embodiment of the present application provides a method for identifying multiple types of crops in irrigation areas by active and passive time-series remote sensing collaboration, including acquiring active remote sensing data and passive remote sensing data; preprocessing the active remote sensing data to obtain SAR remote sensing images; preprocessing the passive remote sensing data to obtain optical remote sensing images; calculating a normalized vegetation index based on the optical remote sensing images; superimposing the SAR remote sensing images, the optical remote sensing images and the normalized vegetation index in chronological order to construct a remote sensing data time-series feature set; constructing an irrigation area multi-type crop sample point set; training multiple base classifiers based on the remote sensing data time-series feature set and the irrigation area multi-type crop sample point set, and then using the multiple base classifiers to respectively identify the multiple types of crops in the irrigation area; and integrating and classifying the recognition results of the multiple base classifiers to obtain a classification result of the irrigation area multi-type crops. This method combines optical remote sensing images and SAR remote sensing images to effectively overcome the problem that effective remote sensing observation data cannot be obtained in irrigation areas due to cloudy and rainy weather; secondly, this method proposes to fuse SAR remote sensing images, optical remote sensing images and normalized vegetation index in chronological order to enhance the expression of crop phenological characteristics and significantly reduce the interference of the "different objects with the same spectrum" phenomenon; finally, this method adopts an integrated classification method to enhance the stability and generalization of multi-type crop identification in irrigation areas. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] The drawings described herein are used to provide further understanding of the present invention and constitute a part of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present application and do not constitute improper limitations on the present invention.

[0048] In the attached picture:

[0049] Figure 1 It is a flow chart of a method for identifying multiple types of crops in irrigation areas by active and passive time-series remote sensing collaboration provided in an embodiment of the present application;

[0050] Figure 2 is a schematic diagram of the geographical location of an irrigation area provided in an embodiment of the present application;

[0051] Figure 3 yes Figure 2 Schematic diagram of SAR remote sensing image of Zhong irrigation area in VV polarization mode;

[0052] Figure 4 yes Figure 2 Schematic diagram of SAR remote sensing image of Zhong irrigation area in VH polarization mode;

[0053] Figure 5 yes Figure 2 Schematic diagram of optical remote sensing images of irrigation areas in;

[0054] Figure 6 yes Figure 2 Schematic diagram of multi-type crop sampling points in the irrigation area;

[0055] Figure 7 yes Figure 2 Schematic diagram of the integrated classification results of multiple types of crops in the irrigation area. DETAILED DESCRIPTION

[0056] In order to enable those skilled in the art to better understand the technical solutions in the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without creative work should fall within the scope of protection of the present application.

[0057] In the description of the embodiments of the present application, the terms "first" and "second" are used for descriptive purposes only and are not to be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one such feature. In the description of the present application, "plurality" means at least two, for example, two, three, etc., unless otherwise clearly and specifically defined.

[0058] In this application, unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, it can be the internal connection of two elements or the interaction relationship between two elements, unless otherwise clearly defined. For ordinary technicians in this field, the specific meanings of the above terms in this application can be understood according to specific circumstances.

[0059] In the present application, unless otherwise clearly specified and limited, a first feature being "above" or "below" a second feature may mean that the first and second features are in direct contact, or the first and second features are in indirect contact through an intermediate medium. Moreover, a first feature being "above", "above" or "above" a second feature may mean that the first feature is directly above or obliquely above the second feature, or simply means that the first feature is higher in level than the second feature. A first feature being "below", "below" or "below" a second feature may mean that the first feature is directly below or obliquely below the second feature, or simply means that the first feature is lower in level than the second feature.

[0060] In modern agricultural production and management, crop types and spatial distribution are important agricultural information. Efficient, accurate and dynamic monitoring of crop information in irrigation areas is of great significance to ensuring food security. Irrigation areas usually have large geographical spans and widely distributed cultivated land. Differences in climate change and field management can easily lead to different time-varying information of the same crop in different plots, and there is a common phenomenon of fragmented plots and diverse crop types.

[0061] Therefore, it is challenging to quickly and accurately obtain crop agricultural information in irrigation areas. Conventional methods mainly rely on manual surveys and statistical reporting, which have shortcomings such as high cost and low efficiency when applied on a large scale, and are difficult to effectively adapt to the emergency decision-making needs of management departments. Remote sensing technology has the characteristics of wide data coverage and strong timeliness, and has become the main technical means for obtaining crop agricultural information. Facing irrigation scenarios, when using remote sensing technology to obtain crop agricultural information on a large scale, it is mainly faced with interference from factors such as cloudy and rainy weather, "different objects with the same spectrum" and "same object with different spectrum", resulting in low accuracy of agricultural information and weak classification generalization.

[0062] Specifically: (1) The scale of irrigation areas is often accompanied by cloudy and rainy weather, especially in the southern region, which makes it difficult to obtain effective optical remote sensing observation data for key crop phenological periods, restricting the large-scale application of optical remote sensing crop extraction; active remote sensing has all-weather and all-day observation capabilities, but is easily interfered by speckle noise and complex scattering characteristics of cultivated areas. (2) The "different objects, same spectrum" phenomenon is widely present in optical remote sensing images, that is, different types of crops have similar spectral characteristics in the same period of imagery, resulting in the inability of classification algorithms to effectively distinguish them, and classification confusion is prone to occur. (3) The "same object, different spectrum" phenomenon often brings difficulties to remote sensing classification tasks, that is, due to local differences in soil environment, farming methods, climate factors, etc., crops of the same type are in different growth stages, resulting in poor classification results of classification algorithms.

[0063] In order to solve the above problems, the embodiment of the present application provides a method and system for identifying multiple types of crops in irrigation areas by active and passive time-series remote sensing collaboration. The solution provided by the embodiment of the present application will be described in detail below in conjunction with the accompanying drawings of the specification.

[0064] Figure 1 It is a flow chart of a method for identifying multiple types of crops in irrigation areas by active and passive time-series remote sensing collaboration provided in an embodiment of the present application; Figure 2 is a schematic diagram of the geographical location of an irrigation area provided in an embodiment of the present application; Figure 3 yes Figure 2 Schematic diagram of SAR remote sensing image of Zhong irrigation area in VV polarization mode; Figure 4 yes Figure 2 Schematic diagram of SAR remote sensing image of Zhong irrigation area in VH polarization mode; Figure 5 yes Figure 2 Schematic diagram of optical remote sensing images of irrigation areas in; Figure 6 yes Figure 2 Schematic diagram of multi-type crop sampling points in the irrigation area; Figure 7 yes Figure 2 Schematic diagram of the integrated classification results of multiple types of crops in the irrigation area.

[0065] Reference Figure 1 As shown, the embodiment of the present application provides a method for identifying multiple types of crops in irrigation areas by active and passive time-series remote sensing collaboration, including:

[0066] S100: Acquire active remote sensing data and passive remote sensing data.

[0067] Specifically, according to the vector range of the irrigation area, the active remote sensing data of the irrigation area and the passive remote sensing data of the irrigation area are obtained. Exemplarily, the active remote sensing data of the irrigation area are Sentinel-1 SAR images, and the passive remote sensing data of the irrigation area are Sentinel-2 optical images. It should be noted that the Sentinel-1 SAR images are SAR images collected by the Sentinel-1 satellite, and the Sentinel-2 optical images are optical images collected by the Sentinel-2 satellite. Sentinel-1 and Sentinel-2 are two sentinel satellites. Sentinel-1 is an Earth observation satellite series under the Copernicus program of the European Space Agency (ESA). The main task of this satellite series is to provide all-weather, all-day radar imaging services for land and ocean observations. The Sentinel-1 satellite is equipped with a C-band synthetic aperture radar (SAR), which can penetrate clouds and darkness to obtain high-resolution radar images; Sentinel-2 is a series of high-resolution multispectral imaging satellites under the European Space Agency's Copernicus program. The main mission of this satellite series is land monitoring, providing images of vegetation, soil and water cover, inland waterways and coastal areas, and can also be used for emergency rescue services. The Sentinel-2 satellite is equipped with a multispectral imager (MSI), which can provide image data in 13 spectral bands, covering visible light, near infrared and short-wave infrared regions, with spatial resolutions of 10 meters, 20 meters and 60 meters respectively, all of which are known to those skilled in the art.

[0068] S200: Preprocessing the active remote sensing data to obtain a SAR remote sensing image.

[0069] Specifically, it includes terrain correction, Refined Lee filtering, stitching, cropping and standard deviation normalization of active remote sensing data to form SAR remote sensing images covering the entire irrigation area from May to October. Among them, the SAR remote sensing image is a multi-view image of the ground distance in the interferometric wide-band mode, the polarization modes are VV polarization and VH polarization, and the image spatial resolution is 10m. Refined Lee filtering is an adaptive filtering algorithm for speckle noise in synthetic aperture radar (SAR) images. It is based on the improvement of classic Lee filtering, which can better retain edge and detail information while suppressing noise. When the polarization mode is VV polarization, the obtained SAR remote sensing image is referenced. Figure 2 As shown in the figure, when the polarization mode is VH polarization, the SAR remote sensing image obtained is referenced Figure 3 shown.

[0070] S300: Preprocessing the passive remote sensing data to obtain an optical remote sensing image.

[0071] Specifically, it uses the QA60 band to remove clouds from passive remote sensing data, and then constructs optical remote sensing images covering the entire irrigation area from May to October through stitching, monthly median synthesis, cropping, and standard deviation normalization.

[0072] The optical remote sensing image is an L2A surface reflectance product. The image used consists of 10 spectral bands, including blue band, green band, red band, vegetation red edge 1 band, vegetation red edge 2 band, vegetation red edge 3 band, near infrared band, narrow edge near infrared band, shortwave infrared 1 band and shortwave infrared 2 band. Among them, the spatial resolution of the blue band, green band, red band and near infrared band is 10m; the spatial resolution of vegetation red edge 1 band, vegetation red edge 2 band, vegetation red edge 3 band, narrow edge near infrared band, shortwave infrared 1 band and shortwave infrared 2 band is 20m. The spatial resolution of the bands with a spatial resolution of 20m needs to be resampled to 10m through nearest neighbor interpolation.

[0073] S400: Calculate the normalized vegetation index based on optical remote sensing images.

[0074] The formula for calculating the normalized difference vegetation index is:

[0075]

[0076] Where NDVI represents the normalized difference vegetation index, ρ R represents the red band reflectance of the image, ρ NIR Represents the image near-infrared band reflectance.

[0077] S500: Superimpose the SAR remote sensing image, the optical remote sensing image and the normalized difference vegetation index in chronological order to construct a time series feature set of remote sensing data.

[0078] Based on the active and passive remote sensing data and vegetation indices in the key phenological period of the irrigation area from May to October, including 12 periods of Sentinel-1SAR images, 6 periods of monthly median synthetic Sentinel-2 optical images and 6 periods of NDVI index, they are stacked in chronological order to form a time series feature set of remote sensing data.

[0079] S600: Construct a sample point set of multiple types of crops in irrigation areas.

[0080] Through the combination of irrigation area survey, high-resolution image identification and irrigation area data, four types of crop samples were obtained, namely corn, rice, soybean and others (lotus root, greenhouse, etc.). For example, the number of sample points is 1110, 978, 188 and 450 respectively. Figure 6 shown.

[0081] S700: Based on the remote sensing data time series feature set and the irrigation area multi-type crop sample point set, multiple base classifiers are trained, and then the multiple base classifiers are used to respectively identify the irrigation area multi-type crops.

[0082] Exemplarily, the remote sensing data time series feature set is respectively input into a random forest base classifier (RF), a rotating forest base classifier (RoF) and a particle swarm optimized support vector machine base classifier (PSO-SVM), and the crop sample points in the irrigation area multi-type crop sample point set are divided into a training set and a test set in a 1:1 ratio. The training set and the test set are used to train and test the random forest base classifier, the rotating forest base classifier and the particle swarm optimized support vector machine base classifier, respectively.

[0083] Among them, RF is a statistical learning method that uses bootstrap resampling theory, which has the characteristics of reliable accuracy and high tolerance to outliers and noise; RoF uses principal component analysis to generate a rotated feature space to improve feature diversity, and performs well in improving classification results; PSO-SVM is a powerful classifier for handling small-scale training samples and nonlinear high-dimensional problems. The number of decision trees in RF and RoF is set to 10, and the number of features in the RF subset uses the default value, that is, the largest integer not greater than the square root of the number of features used.

[0084] The trained multiple base classifiers can identify multiple types of crops in irrigation areas respectively.

[0085] S800: Integrate and classify the recognition results of multiple base classifiers to obtain classification results of multiple types of crops in the irrigation area.

[0086] Specifically, it includes respectively calculating the overall classification accuracy of each base classifier based on the recognition results of multiple base classifiers;

[0087] The recognition results of multiple base classifiers are voted by majority, and the one with the highest number of votes is determined as the classification result of multiple types of crops in the irrigation area. If the votes are the same, the recognition result of the base classifier with the highest overall classification accuracy is determined as the classification result of multiple types of crops in the irrigation area.

[0088] The calculation formula for the overall classification accuracy is:

[0089]

[0090] In the formula, M is the confusion matrix, n is the number of categories, N is the total number of test samples, and M uu Represents the number of samples of category u that are correctly identified as u, u∈[1,n].

[0091] S900: Evaluate the accuracy of the classification results of multiple types of crops in the irrigation area.

[0092] Based on the classification results, the evaluation indicators of overall classification accuracy (OA), average classification accuracy (AA) and Kappa coefficient (κ) were calculated to evaluate the crop recognition ability of the method. In this embodiment, the OA, AA and κ of the integrated classification results of irrigation area crops were 95.43%, 93.87% and 0.93, respectively.

[0093] The calculation formulas of the evaluation indicators OA, AA and κ are as follows:

[0094]

[0095] In the formula, M is the confusion matrix, n is the number of categories, N is the total number of test samples, and M uu Indicates the number of samples of category u that are correctly identified as u, M uv Represents the number of samples of category v that are identified as u, u∈[1,n],v∈[1,n].

[0096] It can be understood that this method combines optical remote sensing images and SAR remote sensing images to effectively overcome the problem that effective remote sensing observation data cannot be obtained in irrigation areas due to cloudy and rainy weather; secondly, this method proposes to fuse SAR remote sensing images, optical remote sensing images and normalized vegetation index in chronological order to enhance the expression of crop phenological characteristics and significantly reduce the interference of the "different objects with the same spectrum" phenomenon; finally, this method adopts an integrated classification method to enhance the stability and generalization of multi-type crop identification in irrigation areas.

[0097] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0098] Based on the same inventive concept, the embodiment of the present application also provides an active and passive time-series remote sensing coordinated irrigation area multi-type crop identification system for implementing the above-mentioned active and passive time-series remote sensing coordinated irrigation area multi-type crop identification method. The implementation scheme for solving the problem provided by the system is similar to the implementation scheme recorded in the above-mentioned method, so the specific limitations in one or more embodiments of the active and passive time-series remote sensing coordinated irrigation area multi-type crop identification system provided below can refer to the above-mentioned limitations on the active and passive time-series remote sensing coordinated irrigation area multi-type crop identification method, which will not be repeated here.

[0099] In the second aspect, an embodiment of the present application provides an irrigation area multi-type crop identification system based on active and passive time-series remote sensing collaboration, which applies the irrigation area multi-type crop identification method based on active and passive time-series remote sensing collaboration as described in the first aspect, including a remote sensing data acquisition module, a preprocessing module, a normalized vegetation index calculation module, a remote sensing data time-series feature set construction module, a sample point set construction module, a crop identification module and an integrated classification module. Among them, the remote sensing data acquisition module is used to acquire active remote sensing data and passive remote sensing data; the preprocessing module is used to preprocess the active remote sensing data to obtain SAR remote sensing images, and is used to preprocess the passive remote sensing data to obtain optical remote sensing images; the normalized vegetation index calculation module is used to calculate the normalized vegetation index based on the optical remote sensing image; the remote sensing data time series feature set construction module is used to superimpose the SAR remote sensing image, the optical remote sensing image and the normalized vegetation index in chronological order to construct the remote sensing data time series feature set; the sample point set construction module is used to construct the irrigation area multi-type crop sample point set; the crop recognition module is used to train multiple base classifiers based on the remote sensing data time series feature set and the irrigation area multi-type crop sample point set, and then use the multiple base classifiers to respectively identify the multiple types of crops in the irrigation area; the integrated classification module is used to integrate and classify the recognition results of multiple base classifiers to obtain the classification results of the irrigation area multi-type crops.

[0100] Each module in the above active and passive time-series remote sensing coordinated irrigation area multi-type crop identification system can be implemented in whole or in part by software, hardware and their combination. The above modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the corresponding operations of each of the above modules.

[0101] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited to this.

[0102] It is easy to understand that those skilled in the art can combine, split, reorganize, etc. the embodiments of the present application to obtain other embodiments based on the several embodiments provided in the present application, and these embodiments do not exceed the protection scope of the present application.

[0103] The above specific implementation methods further explain in detail the purpose, technical solutions and beneficial effects of the embodiments of the present application. It should be understood that the above are only specific implementation methods of the embodiments of the present application and are not used to limit the protection scope of the embodiments of the present application. Any modifications, equivalent substitutions, improvements, etc. made on the basis of the technical solutions of the embodiments of the present application should be included in the protection scope of the embodiments of the present application.

Claims

1. A method for identifying multiple types of crops in irrigation areas based on active and passive time-series remote sensing, characterized in that: include: Acquire active remote sensing data and passive remote sensing data; Preprocessing the active remote sensing data to obtain SAR remote sensing images; Preprocessing the passive remote sensing data to obtain an optical remote sensing image; Calculate the Normalized Difference Vegetation Index based on optical remote sensing images; The SAR remote sensing images, optical remote sensing images and normalized vegetation index are superimposed in chronological order to construct a time series feature set of remote sensing data; Construct a sample point set of multiple types of crops in irrigation areas; Based on the remote sensing data time series feature set and the sample point set of multiple types of crops in the irrigation area, multiple base classifiers are trained, and then the multiple types of crops in the irrigation area are respectively identified using the multiple base classifiers; The recognition results of multiple base classifiers are integrated and classified to obtain the classification results of multiple types of crops in irrigation areas.

2. The method for identifying multiple types of crops in irrigation areas based on active and passive time-series remote sensing collaboration according to claim 1 is characterized in that: The obtaining of active remote sensing data and passive remote sensing data includes: Active remote sensing data and passive remote sensing data are screened out according to the vector range of the irrigation area.

3. The method for identifying multiple types of crops in irrigation areas based on active and passive time-series remote sensing collaboration according to claim 1 is characterized in that: The preprocessing of the active remote sensing data to obtain SAR remote sensing images includes: The active remote sensing data was subjected to terrain correction, Refined Lee filtering, splicing, cropping and standard deviation normalization to form SAR remote sensing images covering the entire irrigation area from May to October; The SAR remote sensing image is a multi-view image of ground distance in an interferometric wide-band mode, the polarization modes are VV polarization and VH polarization, and the image spatial resolution is 10m.

4. The method for identifying multiple types of crops in irrigation areas based on active and passive time-series remote sensing collaboration according to claim 1 is characterized in that: The preprocessing of the passive remote sensing data to obtain an optical remote sensing image includes: The passive remote sensing data were declouded using the QA60 band, and then optical remote sensing images covering the entire irrigation area from May to October were constructed through splicing, monthly median synthesis, cropping, and standard deviation normalization. The optical remote sensing image is an L2A surface reflectance product, and the image used is composed of 10 spectral bands.

5. The method for identifying multiple types of crops in irrigation areas based on active and passive time-series remote sensing collaboration according to claim 1 is characterized in that: The method of calculating the normalized vegetation index based on the optical remote sensing image comprises: The formula for calculating the normalized vegetation index is: Where NDVI represents the normalized difference vegetation index, ρ R represents the red band reflectance of the image, ρ NIR Represents the image near-infrared band reflectance.

6. The method for identifying multiple types of crops in irrigation areas based on active and passive time-series remote sensing collaboration according to claim 1 is characterized in that: The construction of the irrigation area multi-type crop sample point set includes: Crop samples are obtained through a combination of irrigation area surveys, high-resolution image identification, and irrigation area data.

7. The method for identifying multiple types of crops in irrigation areas by active and passive time-series remote sensing collaboration according to claim 1 is characterized in that: The method of training multiple base classifiers based on the remote sensing data time series feature set and the irrigation area multi-type crop sample point set, and then using the multiple base classifiers to respectively identify the irrigation area multi-type crops includes: The remote sensing data time series feature set is respectively input into a random forest base classifier, a rotating forest base classifier and a particle swarm optimized support vector machine base classifier, and the crop sample points in the irrigation area multi-type crop sample point set are divided into a training set and a test set in a 1:1 ratio, and the random forest base classifier, the rotating forest base classifier and the particle swarm optimized support vector machine base classifier are respectively trained and tested using the training set and the test set.

8. The method for identifying multiple types of crops in irrigation areas by active and passive time-series remote sensing collaboration according to claim 1 is characterized in that: The identification results of multiple base classifiers are integrated and classified to obtain the classification results of multiple types of crops in the irrigation area, including: Based on the recognition results of multiple base classifiers, the overall classification accuracy of each base classifier is calculated respectively; The recognition results of multiple base classifiers are voted by majority, and the one with the highest number of votes is determined as the classification result of multiple types of crops in the irrigation area. If the votes are the same, the recognition result of the base classifier with the highest overall classification accuracy is determined as the classification result of multiple types of crops in the irrigation area. The calculation formula for the overall classification accuracy is: In the formula, OA represents the overall classification accuracy, M represents the confusion matrix, n represents the number of categories, N represents the total number of test samples, and M uu Represents the number of samples of category u that are correctly identified as u, u∈[1,n].

9. The method for identifying multiple types of crops in irrigation areas by active and passive time-series remote sensing collaboration according to claim 1 is characterized in that: The active and passive time-series remote sensing coordinated irrigation area multi-type crop identification method also includes: The accuracy of the classification results of multiple types of crops in the irrigation area was evaluated.

10. An active and passive time-series remote sensing coordinated irrigation area multi-type crop identification system, using the active and passive time-series remote sensing coordinated irrigation area multi-type crop identification method as described in any one of claims 1 to 9, characterized in that: include: A remote sensing data acquisition module is used to acquire active remote sensing data and passive remote sensing data; A preprocessing module, used for preprocessing the active remote sensing data to obtain SAR remote sensing images, and for preprocessing the passive remote sensing data to obtain optical remote sensing images; Normalized difference vegetation index calculation module, used to calculate the normalized difference vegetation index based on optical remote sensing images; The remote sensing data temporal feature set construction module is used to superimpose SAR remote sensing images, optical remote sensing images and normalized vegetation index in chronological order to construct the remote sensing data temporal feature set; Sample point set construction module, used to construct sample point sets of multiple types of crops in irrigation areas; A crop identification module is used to train multiple base classifiers based on a remote sensing data time series feature set and a sample point set of multiple types of crops in the irrigation area, and then use the multiple base classifiers to respectively identify the multiple types of crops in the irrigation area; The integrated classification module is used to integrate the recognition results of multiple base classifiers to obtain the classification results of multiple types of crops in the irrigation area.

Images