A method, apparatus, equipment, and storage medium for tea garden identification based on multi-temporal remote sensing images.

Abstract

This application discloses a method, apparatus, equipment, and storage medium for tea garden identification based on multi-temporal remote sensing imagery, relating to the field of agricultural remote sensing technology. The method includes the following steps: determining the phenological stage and agricultural operation period of tea; collecting multi-temporal remote sensing image data corresponding to the area to be identified; calculating the normalized vegetation index (NVI) and a pre-constructed normalized target crop difference index (NCR) based on the remote sensing image data, and obtaining their altitude values ​​and spectral feature values ​​for each band; calculating the weights of the NVI, NCR, altitude features, and spectral features based on the obtained feature values; and classifying the area according to the weight of each feature value to obtain the identification result of the area to be identified. This application fully considers the growth period of tea and the time characteristics of important agricultural operations, and constructs new feature indices based on the remote sensing spectral characteristics of the tea garden during this time period. Combined with a classification algorithm, it can accurately identify tea garden information.

Description

Technical Field

[0001] This application relates to the field of agricultural remote sensing technology, and in particular to a method, apparatus, equipment and storage medium for tea garden identification based on multi-temporal remote sensing images. Background Technology

[0002] With the development of satellite remote sensing imagery, it has become increasingly convenient to acquire high-resolution satellite remote sensing imagery and multispectral and hyperspectral satellite remote sensing imagery data. This has led to an increasing number of technologies for identifying tea garden information using satellite remote sensing imagery. However, traditional identification methods rely on a single spectrum or existing index division methods or a single image, which can easily lead to misclassification or omission. Summary of the Invention

[0003] This application addresses the shortcomings of existing technologies by providing a tea garden identification method based on multi-temporal remote sensing images, taking into account the influence of tea growth period and agricultural operation period on the remote sensing spectrum of tea gardens.

[0004] To achieve the above objectives, this application adopts the following technical solution:

[0005] This application discloses a tea garden identification method based on multi-temporal remote sensing imagery, comprising the following steps:

[0006] Determine the phenological period and agricultural operation period of tea, and collect multi-temporal remote sensing image data of the area to be identified during the phenological period and agricultural operation period;

[0007] The normalized vegetation index and the pre-constructed normalized target crop difference index are calculated based on the remote sensing image data, and the altitude value of the area to be identified and the spectral feature value of each band in the remote sensing image data are obtained.

[0008] The weights of the normalized vegetation index, normalized target crop difference index, altitude feature and spectral feature are calculated based on the obtained feature values.

[0009] The region to be identified is classified according to the weight of each feature value.

[0010] Preferably, the remote sensing image data is provided by the Sentinel-2 satellite. Preferably, after acquiring multi-temporal remote sensing image data of the area to be identified during the phenological and agricultural operation periods, the method further includes:

[0011] The remote sensing image data includes images containing coastal aerosols, water vapor, and cirrus cloud bands, and the filtered remote sensing image data is then subjected to radiometric and atmospheric correction.

[0012] Preferably, the method for constructing the normalized target crop difference index includes:

[0013] Training samples were collected, which included tea garden samples and control group samples.

[0014] The tea garden samples were compared with the control group samples, and the remote sensing image band features corresponding to the tea garden and each control group were extracted respectively.

[0015] A normalized target crop difference index is constructed based on the band characteristics. The normalized target crop difference index is as follows:

[0016] NDCI = (B max -B min ) / (B max +B min )

[0017] Among them, B max B represents the band information that indicates the target crop band has the highest value among all categories at the same time. min This indicates the band information where the target crop band is at its lowest point across all categories at the same time.

[0018] Preferably, the step of calculating the weights of the normalized vegetation index, the normalized target crop difference index, altitude characteristics, and spectral characteristics based on the obtained feature values ​​includes:

[0019] Multiple second classifiers were constructed with tea gardens and non-tea gardens as classification targets, using normalized vegetation index, normalized target crop difference index, altitude characteristics and spectral characteristics as classification features respectively;

[0020] The obtained feature values ​​are input into each of the second classifiers for training, and the sum of the classification accuracy of different categories under the same feature is used as the multi-class weight of the corresponding feature.

[0021] Preferably, the step of inputting the weight of each feature value into a pre-trained first classifier for classification to obtain the recognition result of the region to be identified includes:

[0022] The points that repeatedly appear in different categories and the points of unknown categories in the training results are combined into a new dataset. The new dataset is then input into a pre-trained first classifier that uses the multi-class weights of each feature as the basis for classification, so as to obtain all the recognition results of the region to be identified.

[0023] Preferably, the method further includes: performing temporal phase difference processing on the remote sensing image data to obtain a temporal difference data source, wherein the temporal difference data source is used to characterize the temporal variation features of the remote sensing image.

[0024] A tea garden identification device based on multi-temporal remote sensing imagery includes:

[0025] The acquisition module is used to determine the phenological period and agricultural operation period of tea, and to collect multi-temporal remote sensing image data of the area to be identified during the phenological period and agricultural operation period.

[0026] The extraction module is used to calculate the normalized vegetation index value and the pre-constructed normalized target crop difference index value based on the remote sensing image data, and to obtain the altitude value of the area to be identified and the spectral feature value of each band in the remote sensing image data.

[0027] The partitioning module is used to calculate the weights of the normalized vegetation index, normalized target crop difference index, altitude feature, and spectral feature based on the obtained feature values.

[0028] The recognition module is used to classify the region to be recognized according to the weight of each feature value.

[0029] An electronic device includes a memory and a processor, the memory being used to store one or more computer instructions, wherein the one or more computer instructions are executed by the processor to implement a tea garden identification method based on multi-temporal remote sensing imagery as described in any one of the preceding descriptions.

[0030] A computer-readable storage medium storing a computer program that, when executed by a computer, implements a tea garden identification method based on multi-temporal remote sensing imagery as described in any one of the preceding descriptions.

[0031] This application has the following beneficial effects:

[0032] This application fully considers the growth period of tea and the time characteristics of important agricultural operations, and constructs new feature indices based on the remote sensing spectral characteristics of tea gardens during this period. Combined with classification algorithms, tea garden information can be accurately identified. Attached Figure Description

[0033] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0034] Figure 1 This is a flowchart of a tea garden identification method based on multi-temporal remote sensing images, which is an embodiment of this application.

[0035] Figure 2 This is a schematic diagram of a tea garden identification device based on multi-temporal remote sensing images, which is an embodiment 2 of this application;

[0036] Figure 3This is a schematic diagram of an electronic device that implements a tea garden identification method based on multi-temporal remote sensing images in Embodiment 3 of this application. Detailed Implementation

[0037] The technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0038] The terms “first,” “second,” etc., used in the claims and description of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such terms can be used interchangeably where appropriate. This is merely a way of distinguishing objects with the same attributes in the embodiments of this application. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover non-exclusive inclusion so that a process, method, system, product, or apparatus that comprises a series of units is not necessarily limited to those units, but may include other units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0039] Example 1

[0040] like Figure 1 As shown, a tea garden identification method based on multi-temporal remote sensing imagery includes the following steps:

[0041] S110. Determine the phenological period and agricultural operation period of tea, and collect multi-temporal remote sensing image data of the area to be identified during the phenological period and agricultural operation period.

[0042] S120. Calculate the normalized vegetation index value and the pre-constructed normalized target crop difference index value based on the remote sensing image data, and obtain the altitude value of the area to be identified and the spectral feature value of each band in the remote sensing image data.

[0043] S130. Calculate the weights of the normalized vegetation index, normalized target crop difference index, altitude feature and spectral feature based on the obtained feature values.

[0044] S140. Classify the region to be identified according to the weight of each feature value to obtain the identification result of the region to be identified.

[0045] In Example 1, the phenological period of tea and its corresponding agricultural operation period are first determined. The phenological period refers to the growth, development, and activity patterns of plants and animals and the biological response to the seasons. Here, the emphasis is on tea leaves rather than tea trees. The annual growth and development cycle of tea trees is the physiological, biochemical, and ecological changes of various organs of tea trees in terms of external morphology, internal tissue structure, and internal material composition. It mainly includes the growth and development of tea tree shoots, the development of tea tree roots, the flowering and fruiting of tea trees, and the germination of seeds. This example focuses on the growth and development period of tea tree shoots: In most tea-producing areas of my country, under natural growth conditions, tea trees have three growth and dormancy periods throughout the year: the first growth period is from overwintering bud sprouting to the first growth period, which is the spring shoot period, from about late March to early May, followed by the first dormancy period; the second growth period is the summer shoot period, from about early June to early July, followed by dormancy; the third growth period is the autumn shoot period, which leads to winter dormancy, from about mid-July to early October. However, under manual harvesting conditions, 5 to 6 flushes of new shoots can sprout throughout the year. Therefore, it is necessary to determine the phenological period of tea and the corresponding agricultural operation period for the tea plantation in the area to be identified based on the natural environment and human activities.

[0046] Secondly, remote sensing imagery data was collected as the data to be identified. Based on the tea growing season, starting in March, with a 15-day cycle, and ending at the end of December, satellite remote sensing images with cloud cover of less than 10% were selected as the raw data. The time of the remote sensing images was recorded as T1, T2, ..., T... i , where i is an integer greater than 1, and this is the first data source.

[0047] Furthermore, the remote sensing image data is subjected to temporal phase difference processing to obtain a temporal difference data source, which is used to characterize the temporal variation features of the remote sensing image.

[0048] A key characteristic of tea bushes is pruning. Before pruning, tea bushes appear as green vegetation in images; after pruning, they appear as bare ground. The differences in time-series data can serve as an important data source for identifying tea gardens in this specific situation. Therefore, a temporal difference method is used, i.e., subtracting the previous temporal image from the later temporal image (T). i -T i-1 We obtain i-1 time-varying images, which serve as the second set of data sources—i.e., time-series difference data sources—to characterize this feature of the tea tree.

[0049] If the data source is insufficient, temporal difference can be performed every 2 or 3 time periods to obtain the best recognition image.

[0050] Furthermore, the remote sensing image data was provided by the Sentinel-2 satellite.

[0051] Furthermore, the remote sensing image data includes images containing coastal aerosols, water vapor, and cirrus cloud bands, and the filtered remote sensing image data is then subjected to radiometric and atmospheric correction.

[0052] Next, the bands used for information extraction from all data sources were determined. In this embodiment, the data collected mainly came from the Sentinel-2 satellite, which has a total of 13 bands. Among them, the spatial resolution of the blue, green, red, and near-infrared bands is 10m; the spatial resolution of the three red-edge bands, one narrow-wave near-infrared band, and two short-wave infrared bands is 20m; and the spatial resolution of the coastal aerosol band, water vapor band, and cirrus band is 60m. However, not every band is suitable for extracting information about crops. In particular, the three bands with a spatial resolution of 60m not only have low resolution, but also... The spectral characteristics of these three bands are not closely related to crops, so these three bands can be filtered out. That is, these three bands do not need to be used for crop identification. Only 10 multispectral bands with spatial resolutions of 10m and 20m are selected, namely blue, green, red, red edge 1, red edge 2, red edge 3, near-infrared, narrow-wave near-infrared, short-wave infrared 1 and short-wave infrared 2. Among them, spatial resolution refers to the size of the smallest unit that can be distinguished on the image acquired by remote sensing. It is an indicator used to characterize the details of ground targets that affect resolution. The higher the spatial resolution, the stronger the ability to identify objects. Then, radiometric correction and atmospheric correction are performed on the selected data.

[0053] Then, a tea garden spectral characteristic index, namely the normalized target crop difference index, is constructed. First, the classification categories are determined, which are divided into target crops and non-target crops. In this embodiment, the categories are tea gardens and non-tea gardens, which are also control group crops. The general idea of ​​identification is to analyze the main information of the target crop tea garden among many categories. Specifically, tea belongs to vegetation in the broad category. Among the vegetation categories, the category that distinguishes it from tea crops is selected. Since tea trees are evergreen trees and are similar to shrubs in terms of tree species, and tea gardens have very regular planting characteristics due to agricultural management, forest land and cultivated land categories are the key control groups to be analyzed during identification.

[0054] Furthermore, the normalized target crop difference index is constructed, and the specific steps include:

[0055] Training samples were collected, which included tea garden samples and control group samples.

[0056] The tea garden samples were compared with the control group samples, and the remote sensing image band features corresponding to the tea garden and each control group were extracted respectively.

[0057] A normalized target crop difference index is constructed based on the band characteristics. The normalized target crop difference index is as follows:

[0058] NDCI = (B max -B min ) / (B max +B min )

[0059] Among them, B max B represents the band information that indicates the target crop band has the highest value among all categories at the same time. min This indicates the band information where the target crop band is at its lowest point across all categories at the same time.

[0060] Training samples were collected; in this embodiment, these were farmland, woodland, and tea garden samples, denoted as ROI_Farmland, ROI_Woodland, and ROI_Tea Garden, respectively. Approximately 50-200 training samples were randomly selected for each category. The training samples could be data from different areas acquired by UAVs, and should be as comprehensive as possible. Multi-temporal remote sensing image data was also selected according to phenological periods, with the corresponding remote sensing image times denoted as T1, T2, ..., T... i Then, by analyzing the characteristics of the 10 spectral curves corresponding to these three categories in each temporal image, and based on the information that the target crop, tea, is at the highest and lowest points in the bands among all categories, a normalized target crop difference index is constructed: NDCI = (B max -B min ) / (B max +B min ), where B max B represents the band information that indicates the target crop band has the highest value among all categories at the same time. min This represents the band information where the target crop band is at its lowest point among all categories at the same time. Taking tea garden as an example, NDCI = (B_tea-max - B_tea-min) / (B_tea-max + B_tea-min). B_tea-max represents the band information corresponding to the highest point among all categories at the same time, and B_tea-min represents the band information corresponding to the lowest point among all categories at the same time.

[0061] Simultaneously, the altitude and spectral features corresponding to the tea gardens are extracted from the training samples. Tea generally grows in hilly areas, with an empirical value of over 100m. Then, the Normalized Difference Index of Target Crop and the Normalized Difference Vegetation Index (NDVI) are calculated based on the data to be identified: NDVI = (NIR-R) / (NIR+R), where NIR is the near-infrared band and R is the red band. The NDVI value is between -1 and 1. Tea belongs to vegetation, and an empirical value of 0.2-0.9 is taken. Then, the altitude feature value and various spectral feature values ​​are extracted from the data to be identified.

[0062] Specifically, multiple second classifiers were constructed with tea gardens and non-tea gardens as classification targets, using normalized vegetation index, normalized target crop difference index, altitude characteristics and spectral characteristics as classification features respectively;

[0063] The obtained feature values ​​are input into each of the second classifiers for training, and the sum of the classification accuracy of different categories under the same feature is used as the multi-class weight of the corresponding feature.

[0064] Classification accuracy is the ratio of the number of correct classifications for each category under the same feature to the total number of categories under that category, multiplied by the ratio of the total number of categories under that feature to the total number of categories under that feature. Classification accuracy is used as the single-class weight of the corresponding feature under that category, while the sum of all classification accuracies under that feature is the multi-class weight of that feature. At the same time, this classification may lead to other categories being misclassified into the target category. Therefore, controversial points that appear repeatedly in different categories in the classification results are extracted and combined with points of unknown categories to form a new dataset. At this time, the multi-class weight of each feature, that is, the overall classification contribution of each feature to all categories, is emphasized and the classifier is trained. The category corresponding to the new dataset is determined by voting method, and the final classification result of the region to be identified is obtained by combining them.

[0065] Furthermore, based on actual surveys and land registration data, test samples can be established to analyze the accuracy, Kappa coefficient, and recall rate of classification.

[0066] This embodiment fully considers the growth habits of tea leaves and the impact of agricultural operations on tea gardens. It constructs a characteristic index of the tea garden spectrum by acquiring remote sensing image data within the corresponding time period, and combines it with existing characteristic indices to jointly identify tea gardens in the area to be identified, achieving high accuracy.

[0067] Example 2

[0068] like Figure 2 As shown, a tea garden identification device based on multi-temporal remote sensing imagery includes:

[0069] The acquisition module 10 is used to determine the phenological period and agricultural operation period of tea, and to collect multi-temporal remote sensing image data of the area to be identified during the phenological period and agricultural operation period.

[0070] Extraction module 20 is used to calculate the normalized vegetation index value and the pre-constructed normalized target crop difference index value based on the remote sensing image data, and to obtain the altitude value of the area to be identified and the spectral feature value of each band in the remote sensing image data.

[0071] The partitioning module 30 is used to calculate the weights of the normalized vegetation index, the normalized target crop difference index, the altitude feature, and the spectral feature based on the obtained feature values.

[0072] The recognition module 40 is used to classify the region to be recognized according to the weight of each feature value.

[0073] One embodiment of the above-mentioned device may be as follows: the acquisition module 10 determines the phenological period and agricultural operation period of tea, and collects multi-temporal remote sensing image data of the area to be identified during the phenological period and agricultural operation period; the extraction module 20 calculates the normalized vegetation index (NVI) and the pre-constructed normalized target crop difference index (NDI) based on the remote sensing image data, and obtains the altitude value of the area to be identified and the spectral feature value of each band in the remote sensing image data; the segmentation module 30 calculates the weights of the normalized vegetation index, the normalized target crop difference index, the altitude feature, and the spectral feature based on the obtained feature values; the identification module 40 inputs the weight of each feature value into a pre-trained first classifier for classification, and obtains the identification result of the area to be identified.

[0074] Example 3

[0075] like Figure 3 As shown, an electronic device includes a memory 301 and a processor 302. The memory 301 is used to store one or more computer instructions, wherein the one or more computer instructions are executed by the processor 302 to implement the above-described method for identifying tea gardens based on multi-temporal remote sensing images.

[0076] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working process of the electronic device described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0077] A computer-readable storage medium storing a computer program that, when executed by a computer, enables a tea garden identification method based on multi-temporal remote sensing imagery as described above.

[0078] For example, a computer program can be divided into one or more modules / units. One or more modules / units are stored in memory 301 and executed by processor 302. Data I / O interface transmission is completed by input interface 305 and output interface 306 to complete the present invention. One or more modules / units can be a series of computer program instruction segments capable of performing specific functions. The instruction segments are used to describe the execution process of the computer program in the computer device.

[0079] The computer device can be a desktop computer, laptop, handheld computer, or cloud server, etc. The computer device may include, but is not limited to, a memory 301 and a processor 302. Those skilled in the art will understand that this embodiment is merely an example of a computer device and does not constitute a limitation on the computer device. It may include more or fewer components, or a combination of certain components, or different components. For example, the computer device may also include an input device 307, a network access device, a bus, etc.

[0080] The processor 302 can be a central processing unit (CPU), or other general-purpose processor 302, digital signal processor 302 (DSP), application-specific integrated circuit (ASIC), field-programmable gate array (FPGA), or other programmable logic device, discrete gate or transistor logic device, discrete hardware component, etc. The general-purpose processor 302 can be a microprocessor 302, or any conventional processor 302, etc.

[0081] The memory 301 can be an internal storage unit of a computer device, such as a hard disk or RAM. The memory 301 can also be an external storage device of a computer device, such as a plug-in hard disk, Smart Media Card (SMC), Secure Digital (SD) card, or Flash Card. Furthermore, the memory 301 can include both internal and external storage units. The memory 301 is used to store computer programs and other programs and data required by the computer device. The memory 301 can also be used for temporary storage in the output device 308. The aforementioned storage media include various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM) 303, random access memory (RAM) 304, discs, or optical discs.

[0082] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions within the technical scope disclosed in the present invention should be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A tea garden identification method based on multi-temporal remote sensing imagery, characterized in that, Includes the following steps: Determine the phenological period and agricultural operation period of tea, and collect multi-temporal remote sensing image data of the area to be identified during the phenological period and agricultural operation period; The remote sensing image data containing coastal aerosols, water vapor, and cirrus cloud bands is filtered, and radiometric and atmospheric corrections are performed on the filtered remote sensing image data. The normalized vegetation index and the pre-constructed normalized target crop difference index are calculated based on the remote sensing image data, and the altitude value of the area to be identified and the spectral feature value of each band in the remote sensing image data are obtained. The weights of the normalized vegetation index, normalized target crop difference index, altitude feature, and spectral feature are calculated based on the obtained feature values, including: Multiple second classifiers were constructed with tea gardens and non-tea gardens as classification targets, using normalized vegetation index, normalized target crop difference index, altitude characteristics and spectral characteristics as classification features respectively; Each obtained feature value is input into each of the second classifiers for training, and the sum of the classification accuracy of different categories under the same feature is used as the multi-class weight of the corresponding feature. The classification is performed based on the weight of each feature value to obtain the recognition result of the region to be identified, including: The points that repeatedly appear in different categories and the points of unknown categories in the training results are combined into a new dataset. Voting is performed based on the multi-class weight of each feature to obtain all the recognition results of the region to be identified. The method for constructing the normalized target crop difference index includes: Training samples were collected, which included tea garden samples and control group samples. The tea garden samples were compared with the control group samples, and the remote sensing image band features corresponding to the tea garden and each control group were extracted respectively. A normalized target crop difference index is constructed based on the band characteristics. The normalized target crop difference index is as follows: NDCI=(B max -B min ) / (B max +B min ) Among them, B max B represents the band information that indicates the target crop band has the highest value among all categories at the same time. min This indicates the band information where the target crop band is at its lowest point across all categories at the same time.

2. The tea garden identification method based on multi-temporal remote sensing imagery according to claim 1, characterized in that, The remote sensing image data was provided by the Sentinel-2 satellite.

3. The tea garden identification method based on multi-temporal remote sensing imagery according to claim 1, characterized in that, The method further includes: performing temporal phase difference processing on the remote sensing image data to obtain a temporal difference data source, wherein the temporal difference data source is used to characterize the temporal variation features of the remote sensing image.

4. A tea garden identification device based on multi-temporal remote sensing imagery, characterized in that, include: The acquisition module is used to determine the phenological period and agricultural operation period of tea, and to collect multi-temporal remote sensing image data of the area to be identified during the phenological period and agricultural operation period. The extraction module is used to calculate the normalized vegetation index value and the pre-constructed normalized target crop difference index value based on the remote sensing image data, and to obtain the altitude value of the area to be identified and the spectral feature value of each band in the remote sensing image data. The partitioning module is used to calculate the weights of the normalized vegetation index, normalized target crop difference index, altitude feature, and spectral feature based on the obtained feature values, including: Multiple second classifiers were constructed with tea gardens and non-tea gardens as classification targets, using normalized vegetation index, normalized target crop difference index, altitude characteristics and spectral characteristics as classification features respectively; Each obtained feature value is input into each of the second classifiers for training, and the sum of the classification accuracy of different categories under the same feature is used as the multi-class weight of the corresponding feature. The recognition module is used to classify the region to be recognized according to the weight of each feature value, and obtain the recognition result of the region to be recognized, including: The points that repeatedly appear in different categories and the points of unknown categories in the training results are combined into a new dataset. Voting is performed based on the multi-class weight of each feature to obtain all the recognition results of the region to be identified. The method for constructing the normalized target crop difference index includes: Training samples were collected, which included tea garden samples and control group samples. The tea garden samples were compared with the control group samples, and the remote sensing image band features corresponding to the tea garden and each control group were extracted respectively. A normalized target crop difference index is constructed based on the band characteristics. The normalized target crop difference index is as follows: NDCI=(B max -B min ) / (B max +B min ) Among them, B max B represents the band information that indicates the target crop band has the highest value among all categories at the same time. min This indicates the band information where the target crop band is at its lowest point across all categories at the same time. After collecting multi-temporal remote sensing image data of the area to be identified during the phenological and agricultural operation periods, the process also includes: The remote sensing image data includes images containing coastal aerosols, water vapor, and cirrus cloud bands, and the filtered remote sensing image data is then subjected to radiometric and atmospheric correction.

5. An electronic device, characterized in that, The system includes a memory and a processor, wherein the memory is used to store one or more computer instructions, wherein the one or more computer instructions are executed by the processor to implement a tea garden identification method based on multi-temporal remote sensing imagery as described in any one of claims 1 to 3.

6. A computer-readable storage medium storing a computer program, characterized in that, The computer program enables the computer to implement a tea garden identification method based on multi-temporal remote sensing images as described in any one of claims 1 to 3.

Images