A method and system for identifying a farmland irrigation event by SAR remote sensing with anti-rainfall interference

Abstract

The application discloses an anti-rainfall interference farmland irrigation event SAR remote sensing identification method and system, and relates to the field of agricultural remote sensing. The method comprises the following steps: acquiring remote sensing data of a plurality of pixels in a target irrigation area and land types corresponding to the plurality of pixels; determining first change values to fourth change values, a first relative change value and a second relative change value according to VV polarization data of a target farmland area and a target vegetation area in continuous first, second and third time phases; determining a significant water supplement result of the target farmland area according to the first change value and a preset water supplement threshold value; and when the target farmland area has a significant water supplement, determining an irrigation result of the target farmland area according to the second change value, the preset water supplement threshold value, the first relative change value, a first preset significant difference threshold value, the second relative change value and a second preset significant difference threshold value. The application can identify irrigation events without relying on meteorological observation data and has the characteristic of anti-rainfall interference.

Description

Technical Field

[0001] This application relates to the field of agricultural remote sensing technology, and in particular to a SAR remote sensing identification method and system for farmland irrigation events that are resistant to precipitation interference. Background Technology

[0002] Irrigation and precipitation events can both increase soil moisture in farmland, which in turn leads to an increase in the backscattering coefficient of SAR (Synthetic Aperture Radar) VV polarization (Vertical transmit-Vertical receive).

[0003] To distinguish between irrigation events and precipitation events, existing methods generally adopt the approach of "spatial comparison between cultivated land and reference area (natural vegetation area): the backscattering changes of cultivated land and reference area are observed simultaneously. If only the cultivated land area shows a significant increase while the reference area does not change significantly, it is determined to be an irrigation event; if both increase simultaneously, it is determined to be a precipitation event.

[0004] However, existing methods still suffer from systematic missed detections in scenarios involving both heavy rainfall and irrigation, as well as light rainfall and irrigation, failing to determine whether an irrigation event has occurred. To address this issue, current methods must incorporate meteorological observation data (such as precipitation) as an auxiliary criterion. When precipitation exceeds a certain threshold (e.g., 6 mm), it is considered a valid precipitation event, excluding irrigation. However, meteorological observation data suffers from uneven station distribution, limited spatial coverage, and data acquisition delays, making it impossible to accurately determine whether an irrigation event has occurred in areas with sparse meteorological stations.

[0005] Therefore, there is an urgent need for a SAR remote sensing method for identifying farmland irrigation events that is resistant to precipitation interference, so as to identify irrigation events without relying on meteorological observation data. Summary of the Invention

[0006] The purpose of this application is to provide a SAR remote sensing identification method and system for farmland irrigation events that is resistant to precipitation interference, so as to identify irrigation events without relying on meteorological observation data and improve the identification effect under precipitation interference.

[0007] To achieve the above objectives, this application provides the following solution: Firstly, this application provides a SAR remote sensing identification method for farmland irrigation events that is resistant to precipitation interference, including: Acquire remote sensing data of multiple pixels in the target irrigation area, as well as the land types corresponding to multiple pixels. The remote sensing data includes VV polarization data, and the land types include cultivated areas and vegetated areas. Based on the VV polarization data of the target cultivated area and the target vegetation area in the first, second, and third consecutive time phases, the first change value, the second change value, the third change value, the fourth change value, the first relative change value, and the second relative change value are determined. The first change value is the change in VV polarization data of the target cultivated area from the first time phase to the second time phase; the second change value is the change in VV polarization data of the target vegetation area from the first time phase to the second time phase; the third change value is the change in VV polarization data of the target cultivated area from the first time phase to the third time phase; and the fourth change value is the change in VV polarization data of the target vegetation area from the first time phase to the third time phase. The first relative change value is the difference between the first change value and the second change value; and the second relative change value is the difference between the third change value and the fourth change value. Based on the first change value and the preset water replenishment threshold, the significant water replenishment results for the target cultivated area are determined; When significant water replenishment occurs in the target cultivated area, the irrigation result of the target cultivated area is determined based on the second change value, the preset water replenishment threshold, the first relative change value, the first preset significant difference threshold, the second relative change value, and the second preset significant difference threshold. Determining the irrigation results of the target cultivated area includes: when the second change value is greater than the preset water replenishment threshold, determining whether an irrigation event has occurred in the target cultivated area from the first time phase to the third time phase based on the second relative change value and the second preset significant difference threshold.

[0008] Optionally, the determination of the land type corresponding to multiple pixels includes: Based on the historical NDVI data of multiple pixels, determine the maximum NDVI data of multiple pixels in the first preset period before the crop harvest period, and the minimum NDVI data of multiple pixels in the crop harvest period. The harvest index is determined based on the maximum and minimum NDVI data. Based on the harvest index, determine the first and second division thresholds; Based on the harvest index, the first classification threshold, and the second classification threshold, the land type corresponding to multiple pixels is determined.

[0009] Optionally, this SAR remote sensing identification method for farmland irrigation events that is resistant to precipitation interference also includes: When no significant water replenishment occurs in the target cultivated area, the fifth change value is determined based on the VV polarization data of the target cultivated area in the second and third time phases. The irrigation results for the target cultivated area are determined based on the first change value, the fifth change value, the third change value, the preset water replenishment threshold, the second relative change value, and the second preset significant difference threshold.

[0010] Optionally, after determining the significant water replenishment results for the target cultivated area based on the first change value and a preset water replenishment threshold, this SAR remote sensing identification method for farmland irrigation events resistant to precipitation interference further includes: The second preset significant difference threshold is updated based on the duration of the first to third time phases and the first preset significant difference threshold.

[0011] Optionally, the second preset significant difference threshold is updated based on the duration from the first time phase to the third time phase and the first preset significant difference threshold, including: The adjustment coefficient is determined based on the duration of the first to third time phases and the duration of the reference three time phases; The second preset significant difference threshold is determined based on the adjustment coefficient and the first preset significant difference threshold.

[0012] Optionally, the remote sensing data also includes VH polarization data and NDVI data. This precipitation-resistant SAR remote sensing identification method for farmland irrigation events further includes: Based on the VH polarization data of the target cultivated area and the target vegetation area in the first and second time phases, respectively, and the change values ​​of NDVI data of the target cultivated area from the first to the second time phase, the results of agricultural activities in the target cultivated area are determined.

[0013] Secondly, this application provides a SAR remote sensing identification system for farmland irrigation events that is resistant to precipitation interference, comprising: The acquisition module is used to acquire remote sensing data of multiple pixels in the target irrigation area, as well as the land types corresponding to the multiple pixels. The remote sensing data includes VV polarization data, and the land types include cultivated areas and vegetated areas. The calculation module is used to determine the first change value, the second change value, the third change value, the fourth change value, the first relative change value, and the second relative change value based on the VV polarization data of the target cultivated area and the target vegetation area in the first, second, and third consecutive time phases, respectively. The first change value is the change in VV polarization data of the target cultivated area from the first time phase to the second time phase; the second change value is the change in VV polarization data of the target vegetation area from the first time phase to the second time phase; the third change value is the change in VV polarization data of the target cultivated area from the first time phase to the third time phase; and the fourth change value is the change in VV polarization data of the target vegetation area from the first time phase to the third time phase. The first relative change value is the difference between the first change value and the second change value; and the second relative change value is the difference between the third change value and the fourth change value. The determination module is used to determine the significant water replenishment result of the target cultivated area based on the first change value and the preset water replenishment threshold; when significant water replenishment occurs in the target cultivated area, the irrigation result of the target cultivated area is determined based on the second change value, the preset water replenishment threshold, the first relative change value, the first preset significant difference threshold, the second relative change value, and the second preset significant difference threshold. Determining the irrigation results of the target cultivated area includes: when the second change value is greater than the preset water replenishment threshold, determining whether an irrigation event has occurred in the target cultivated area from the first time phase to the third time phase based on the second relative change value and the second preset significant difference threshold.

[0014] Thirdly, this application provides a computer device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the SAR remote sensing identification method for farmland irrigation events with resistance to precipitation interference as described above.

[0015] Fourthly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-mentioned method for SAR remote sensing identification of farmland irrigation events resistant to precipitation interference.

[0016] Fifthly, this application provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the SAR remote sensing identification method for farmland irrigation events resistant to precipitation interference as described above.

[0017] According to the specific embodiments provided in this application, the following technical effects are disclosed: This application provides a SAR remote sensing identification method for farmland irrigation events that is resistant to precipitation interference. By introducing VV polarization data from three consecutive time phases, a dual relative change value criterion is constructed for cultivated areas and vegetated areas. The method utilizes the essential physical difference that irrigation events keep cultivated areas relatively moist while precipitation events cause the humidity of both areas to decrease synchronously. When the humidity of both areas increases synchronously in the dual time phase (from the first time phase to the second time phase), the second relative change value is used to verify whether the moist state of cultivated areas relative to vegetated areas continues. This enables irrigation identification in scenarios where traditional dual time phase methods are prone to missing detection, such as irrigation with heavy precipitation and irrigation with light precipitation. Meanwhile, the irrigation event determination process in this application only uses VV polarization data and three thresholds, without relying on meteorological observation data, and without complex parameter calibration or machine learning training, which improves the efficiency of event determination and environmental adaptability (it can still be applied in areas with sparse meteorological stations or no meteorological records). Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments 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.

[0019] Figure 1An application environment diagram for a SAR remote sensing identification method for farmland irrigation events that is resistant to precipitation interference, provided in an embodiment of this application; Figure 2 A flowchart illustrating a SAR remote sensing identification method for farmland irrigation events resistant to precipitation interference, provided in an embodiment of this application; Figure 3 A grid distribution map of irrigation frequency during the growing season provided in this application embodiment; Figure 4 A schematic diagram of the functional modules of a SAR remote sensing identification system for farmland irrigation events that is resistant to precipitation interference, provided in an embodiment of this application; Figure 5 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0020] The technical solutions of 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.

[0021] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0022] Irrigation and precipitation events can both increase soil moisture in farmland, which in turn leads to an increase in the backscattering coefficient of SAR (Synthetic Aperture Radar) VV polarization (Vertical transmit-Vertical receive).

[0023] To distinguish between irrigation events and precipitation events, existing methods generally adopt the approach of "spatial comparison between cultivated land and reference area (natural vegetation area): the backscattering changes of cultivated land and reference area are observed simultaneously. If only the cultivated land area shows a significant increase while the reference area does not change significantly, it is determined to be an irrigation event; if both increase simultaneously, it is determined to be a precipitation event.

[0024] However, existing methods still produce systematic missed detections in both "heavy rainfall superimposed on irrigation" and "light rainfall superimposed on irrigation" scenarios, making it impossible to determine whether an irrigation event has occurred.

[0025] "Heavy rainfall superimposed on irrigation" scenario: When a heavy rainfall event occurs with a volume equivalent to irrigation water, both the cultivated area and the reference area will rise significantly at the same time, and the spatial control index will be close to zero; if irrigation is actually superimposed during this period, the existing method will still judge it as a simple rainfall event and miss the irrigation event. "Small precipitation superimposed on irrigation" scenario: When a small precipitation event occurs at a certain moment (e.g., on the day a satellite passes overhead), the backscattering coefficient of the reference area increases due to short-term soil surface wetting or vegetation canopy interception response. At the same time, if the cultivated area is already in a high humidity state due to irrigation, both are at a high humidity level and the changes in backscattering coefficients are similar. The spatial control indicators are also not significant. This kind of "small precipitation superimposed on irrigation" scenario will also be misjudged as a precipitation event by existing methods, resulting in missed detection of irrigation events.

[0026] To address this missed detection problem, existing methods must incorporate meteorological observation data (such as precipitation) as an auxiliary criterion. When precipitation exceeds a certain threshold (such as 6 mm), it is considered a valid precipitation event, excluding irrigation. However, meteorological observation data suffers from uneven station distribution, limited spatial coverage, and data acquisition lag. Therefore, in areas with sparse meteorological stations, it is impossible to accurately determine whether an irrigation event has occurred.

[0027] Therefore, there is an urgent need for a SAR remote sensing method for identifying farmland irrigation events that is resistant to precipitation interference, so as to identify irrigation events without relying on meteorological observation data.

[0028] The SAR remote sensing identification method for farmland irrigation events with resistance to precipitation interference provided in this application embodiment can be applied to, for example... Figure 1 In the application environment shown, terminal 102 communicates with server 104 via a network. A data storage system can store the data that server 104 needs to process. The data storage system can be set up independently, integrated into server 104, or placed in the cloud or on another server. Terminal 102 can send the acquired pixel remote sensing data and land type to server 104. After receiving the pixel remote sensing data and land type, server 104 determines multiple change values ​​and change value differences based on the VV polarization data of the target cultivated area and target vegetation area in the first, second, and third consecutive time phases, respectively. Based on the change values ​​and change differences, server 104 determines whether an irrigation event has occurred. Server 104 can feed back the determined irrigation result to terminal 102. Furthermore, in some embodiments, the SAR remote sensing identification method for farmland irrigation events resistant to precipitation interference can also be implemented separately by the server 104 or the terminal 102. For example, the terminal 102 can directly process the remote sensing data and land type of the acquired pixels, or the server 104 can obtain the remote sensing data and land type of the pixels from the data storage system and process the remote sensing data and land type of the acquired pixels.

[0029] The terminal 102 can be, but is not limited to, various desktop computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. IoT devices can include smart speakers, smart TVs, smart air conditioners, and smart in-vehicle devices. Portable wearable devices can include smartwatches, smart bracelets, and head-mounted devices. The server 104 can be implemented using a standalone server or a server cluster composed of multiple servers, or it can be a cloud server.

[0030] In one exemplary embodiment, such as Figure 2 As shown, a SAR remote sensing identification method for farmland irrigation events with resistance to precipitation interference is provided. This method is executed by computer equipment, specifically by a terminal or server alone, or by both a terminal and a server. In this embodiment, the method is applied to... Figure 1 Taking server 104 as an example, the following steps are taken: S201 to S204.

[0031] S201. Obtain remote sensing data of multiple pixels in the target irrigation area, as well as the land types corresponding to the multiple pixels.

[0032] The remote sensing data includes VV polarization data, and the land types include cultivated areas and vegetated areas.

[0033] For example, the target irrigation district is a specific agricultural irrigation area that requires monitoring of irrigation events, statistics of actual irrigated area, and management of irrigation water. It is an area designated by the water conservancy department and has fixed irrigation facilities (such as water diversion canals and pumping stations) and irrigation management units.

[0034] It is understandable that the meteorological conditions such as precipitation and evapotranspiration in the target irrigation area are relatively uniform, belonging to a unified climate zone.

[0035] It can be understood that VV polarization data is the backscattering coefficient data of SAR image data from remote sensing satellites after VV polarization; SAR image data has all-weather, all-time observation capability, is not affected by cloud and rain weather, and can stably acquire backscattering information of the ground surface.

[0036] It can be understood that multiple pixels in the target irrigation area are the same pixels in the SAR image data; the pixels correspond to the actual land.

[0037] For example, the land types corresponding to multiple pixels can be pre-defined by humans or determined by classification based on remote sensing data, without any limitation.

[0038] Existing general-purpose land use products (such as global products like ESRI Land Cover based on deep learning) tend to misclassify vegetation such as woodland, grassland, and riverbanks within and around irrigation areas as arable land in crop rotation zones. This is because crop rotation zones have diverse crop types (e.g., the target irrigation area simultaneously grows winter wheat, rapeseed, and garlic), and the spectral characteristics of different crops vary significantly. Furthermore, the spectral characteristics of the same plot at different seasons are similar to the seasonal changes of natural vegetation at certain times, making it difficult for classification methods based on static spectral characteristics to distinguish them.

[0039] In some possible implementations, the determination of land types corresponding to multiple pixels includes: Based on the historical NDVI data of multiple pixels, determine the maximum NDVI data of multiple pixels in the first preset period before the crop harvest period, and the minimum NDVI data of multiple pixels in the crop harvest period. The harvest index is determined based on the maximum and minimum NDVI data. Based on the harvest index, determine the first and second division thresholds; Based on the harvest index, the first classification threshold, and the second classification threshold, the land type corresponding to multiple pixels is determined.

[0040] For example, historical NDVI (Normalized Difference Vegetation Index) data of multiple pixels can be calculated from the optical remote sensing data (multispectral effects) of the target irrigation area, which is a conventional technique in this field and will not be described in detail here.

[0041] For example, the first preset period before the harvest period can be the vigorous growth period of the crop 1 to 2 months before harvest, and the harvest period can be the month when the crop is harvested.

[0042] For example, the difference between the maximum NDVI data and the minimum NDVI data can be calculated, and the result of the difference can be determined as the harvest index.

[0043] It is understandable that the physical meaning of this harvest index is that: when crops on cultivated land are removed by humans during harvest, the vegetation cover drops sharply from a high value (generally NDVI > 0.4) to a low value (generally NDVI < 0.2) in a short period of time (usually 1-2 weeks), resulting in a large difference; while natural vegetation (woodland, grassland, etc.) does not have human harvesting behavior during the growing season, and its NDVI changes slowly during the crop harvest period, with a lower difference.

[0044] For example, the multi-level Otsu method can be used to divide the harvest index into thresholds with a classification number of 3, resulting in a first division threshold and a second division threshold.

[0045] It should be noted that the multi-level Otsu method is a conventional existing technique and will not be elaborated upon here; it is understood that other methods can also be used to determine the division threshold, and there are no restrictions on this.

[0046] For example, taking a first division threshold as a smaller threshold and a second division threshold as a larger threshold, pixels with a harvest index greater than the second division threshold can be identified as major crop areas, pixels with a harvest index less than or equal to the second division threshold but greater than the first division threshold can be identified as other crop areas, pixels with a harvest index less than or equal to the first division threshold can be identified as vegetation areas, and major crop areas and other crop areas can be identified as cultivated areas.

[0047] It should be noted that determining land type based on the harvest index has the following technical advantages: It is not sensitive to crop type: it can be identified as long as there is harvesting behavior, whether it is winter wheat, rapeseed or garlic, and is suitable for mixed crop planting in crop rotation areas; Insensitive to interannual variation: Harvesting behavior is not affected by interannual variations in precipitation, avoiding the problem of misclassifying rain-fed vegetation as irrigated crops in years with abundant rainfall, which is based on differences in vegetation indices during the growing season. No training samples required: This avoids the limitation of supervised classification methods that require recollecting training data in different regions.

[0048] In this way, instead of relying on the static spectral characteristics of crops during their growth period, the unique harvest events of cultivated land can be used as the basis for identification, which helps to improve the reliability of pixel-based land type identification.

[0049] The technical effects of this embodiment are illustrated below with a specific example.

[0050] Taking a target irrigation area in Puyang City, Henan Province as an example, the maximum NDVI value in May 2025 (the peak crop growth period) and the minimum NDVI value in June 2025 (the harvest period) were calculated to obtain the HI. The multi-level Otsu method was applied to automatically determine two thresholds: the first threshold was 0.2637 and the second threshold was 0.5547. The classification results were: main crop (winter wheat) HI > 0.5547, other crops 0.2637 < HI ≤ 0.5547, and natural vegetation (reference area) HI ≤ 0.2637. Evaluation using 300 ground validation sampling points showed an overall accuracy of 98%, an F-score of 0.9875, and a Kappa coefficient of 0.9375.

[0051] The farmland identification results of this invention were compared with the ESRI Land Cover product (the ESRI product uses a deep learning AI land classification model trained on 400,000 Earth observation images, with a spatial resolution of 10 meters). The comparison results show that the ESRI product had a misclassification error of 10.12%, a missed classification error of 3.75%, an overall accuracy of 88%, and a Kappa coefficient of 0.59; while the present invention had a misclassification error of only 1.25%, a missed classification error of 1.25%, an overall accuracy of 98%, and a Kappa coefficient of 0.938. The main reason for the lower accuracy of the ESRI product is that it incorrectly classifies natural vegetation such as woodland and grassland within and around the irrigation area as farmland.

[0052] For example, after acquiring remote sensing data, the remote sensing data can be preprocessed, including orbit correction, thermal noise removal, radiometric calibration, terrain correction, and speckle noise filtering.

[0053] Among them, the speckle noise filtering can use a mean filter, and the filtering window size can be 5×5 pixels.

[0054] The formula for calculating mean filtering is: in, is the backscattering coefficient (i.e., VV polarization data, in dB) corresponding to the filtered pixel; N is the total number of pixels in the filtering window (5×5=25 in this embodiment). For the pixels within the filtering window The original backscattering coefficient, , These represent the rows and columns of the filter window, respectively.

[0055] It is understandable that speckle noise is an inherent characteristic of SAR coherent imaging, leading to random fluctuations in the backscattering coefficient. Preprocessing to eliminate systematic errors and random noise in SAR data can improve the measurement accuracy of the backscattering coefficient. Experimental results show that after applying this filter, the correlation (R) between the VV polarization backscattering coefficient and soil moisture in the pixel of the Fanxian observation station increased from 0.466 to 0.566, an improvement of 21.4%.

[0056] S202. Based on the VV polarization data of the target cultivated area and the target vegetation area in the first, second, and third consecutive time phases, respectively, determine the first change value, the second change value, the third change value, the fourth change value, the first relative change value, and the second relative change value.

[0057] The first change value is the change value of VV polarization data from the first time phase to the second time phase in the target cultivated area; the second change value is the change value of VV polarization data from the first time phase to the second time phase in the target vegetation area; the third change value is the change value of VV polarization data from the first time phase to the third time phase in the target cultivated area; and the fourth change value is the change value of VV polarization data from the first time phase to the third time phase in the target vegetation area. The first relative change value is the difference between the first change value and the second change value; and the second relative change value is the difference between the third change value and the fourth change value.

[0058] For example, the target cultivated area can be any cell of the land category of cultivated area, and the target vegetation area is the vegetation area located in the same climate unit as the target cultivated area and within the spatial neighborhood (e.g., 5 to 10 km). Both can respond synchronously to local and regional precipitation events.

[0059] It is understandable that the first, second, and third time phases are three consecutive satellite observation time phases, with the time interval between any two adjacent time phases being consistent, and there is no restriction on the size of the time interval between any two adjacent time phases.

[0060] For example, the difference between the second-phase VV polarization data and the first-phase VV polarization data of the target cultivated area can be calculated to obtain a first change value C12; the difference between the second-phase VV polarization data and the first-phase VV polarization data of the target vegetation area can be calculated to obtain a second change value N12; the difference between the third-phase VV polarization data and the first-phase VV polarization data of the target cultivated area can be calculated to obtain a third change value C13; the difference between the third-phase VV polarization data and the first-phase VV polarization data of the target vegetation area can be calculated to obtain a fourth change value N13; the difference between the first change value and the second change value can be calculated to obtain a first relative change value S12; and the difference between the third change value and the fourth change value can be calculated to obtain a second relative change value S13.

[0061] It should be noted that the decision structure of this application is based on the following core physical observations: Irrigation, as a water replenishment event with a large volume of water, shows a significant positive difference between the cultivated area and the reference area that is maintained for a long time; general precipitation (i.e., non-effective precipitation events with small amounts of precipitation), as an instantaneous wetting event, causes the difference between the cultivated area and the reference area to rapidly return to zero after the event ends. The two differ fundamentally in the temporal duration dimension of the signal, which is the physical basis for the decision of this invention.

[0062] Specifically, irrigation is a localized human-induced water replenishment event with a large amount of water replenishment and a long duration, resulting in a persistently wet state in cultivated areas relative to the reference area. The difference between cultivated areas and the reference area can remain significantly positive over multiple satellite observation cycles. Precipitation is a regional meteorological event that affects both cultivated areas and the reference area. Moreover, the retention time of precipitation (especially ineffective small precipitation) in soil moisture is much shorter than that of irrigation. Soil moisture in cultivated areas and the reference area decreases synchronously over time, and the difference between cultivated areas and the reference area returns to zero after the precipitation ends.

[0063] Existing irrigation identification methods employ a dual-temporal spatial comparison decision, calculating S12 based solely on the first and second temporal phases and comparing it with a threshold. This method introduces systematic errors in three scenarios: missed detection of precipitation-irrigation composite events, missed detection of irrigation events superimposed with light precipitation, and missed detection of progressive irrigation, as detailed in the following analysis. To address these errors, existing methods must incorporate external meteorological observation data as auxiliary criteria, but this is limited by the distribution of meteorological stations and the timeliness of the data.

[0064] S203. Based on the first change value and the preset water replenishment threshold, determine the significant water replenishment results for the target cultivated area.

[0065] It is understandable that irrigation and precipitation both lead to increased soil moisture, which in turn causes an increase in the backscattering coefficient.

[0066] For example, when the first change value C12 is greater than the preset water replenishment threshold T1, it can be determined that a significant water replenishment event has occurred in the target cultivated area between the first and second time phases, that is, an irrigation event or a precipitation event may have occurred; when the first change value C12 is less than or equal to the preset water replenishment threshold T1 and is greater than zero, it indicates that the direction of VV polarization change in the cultivated area is positive but the magnitude is not significant, and it can be determined that no significant water replenishment event has occurred.

[0067] It is understandable that when the first change value C12 is less than zero, it indicates that the direction of VV polarization change in the cultivated area is negative, that is, the cultivated area is becoming increasingly dry, which can be determined as a water-free event; this application will not explore this further.

[0068] For example, the preset water replenishment threshold can be 1.0 dB.

[0069] S204. When significant water replenishment occurs in the target cultivated area, the irrigation result of the target cultivated area is determined based on the second change value, the preset water replenishment threshold, the first relative change value, the first preset significant difference threshold, the second relative change value, and the second preset significant difference threshold.

[0070] The determination of irrigation results for the target cultivated area includes: when the second change value is greater than the preset water replenishment threshold, determining whether an irrigation event has occurred in the target cultivated area from the first time phase to the third time phase based on the second relative change value and the second preset significant difference threshold.

[0071] For example, when a significant water replenishment occurs in the target cultivated area (i.e., the first change value C12 is greater than the preset water replenishment threshold T1), the second change value N12 is less than or equal to the preset water replenishment threshold T1, and the first relative change value S12 is greater than the first preset significant difference threshold T2, it can be determined that an irrigation event has occurred in the target cultivated area from the first time phase to the second time phase. Understandably, in this scenario, the VV polarization data of the target cultivated area increases significantly, while the reference area (i.e., the target vegetation area) does not increase synchronously, and the difference between the cultivated area and the reference area is significant, which is consistent with the judgment logic of irrigation events. Therefore, it can be determined that an irrigation event occurred in the target cultivated area between the first and second time phases.

[0072] It should be noted that when significant water replenishment occurs in the target cultivated area and the second change value exceeds the preset water replenishment threshold, there are two possibilities, including scenario A and scenario B: Scenario A: A single precipitation event occurred between the first and second time phases, affecting both the cultivated area and the reference area simultaneously. Precipitation includes heavy precipitation (effective precipitation, where both areas (cultivated area and reference area) show significant increases (i.e., both the first change value C12 and the second change value N12 are greater than the preset water replenishment threshold)) and light precipitation (ineffective precipitation, which, if it occurs on the day of satellite transit, briefly moistens the soil surface, and although it may cause both the first change value C12 and the second change value N12 to exceed the preset water replenishment threshold, the amount of water is insufficient to maintain prolonged moisture). Under both types of precipitation, soil moisture in both areas decreases synchronously over time, thus S13 approaches zero; Scenario B involves simultaneous precipitation and irrigation occurring between the first and second time phases, which can be further divided into two sub-scenarios: B1. Heavy rainfall superimposed with irrigation: Rainfall significantly increased the VV polarization data in the reference area, and irrigation significantly increased the VV polarization data in the cultivated area. The VV polarization data in both areas increased synchronously from the first time phase to the second time phase, or the increase in the cultivated area was slightly greater. B2. Light precipitation superimposed with irrigation: Light precipitation in the second phase increases the VV of the reference area, while the cultivated area is already in a high humidity state due to irrigation. Both are at a high humidity level, and the variation range of VV polarization data is similar, with S12 not significant. This situation is almost indistinguishable from the simple light precipitation in situation A under dual-phase observation.

[0073] Although it is difficult to distinguish between scenario A and scenario B under dual-time-phase observation (first and second time-phase), they can show essential differences under the third time-phase: under scenario A, the humidity in both areas decreases synchronously, and S13 is close to zero; under scenario B, the cultivated area remains relatively humid compared to the reference area due to the large amount of irrigation water, and S13 is a significantly positive value (i.e., the second relative change value is greater than the second preset significant difference threshold).

[0074] Based on this difference in circumstances, when significant water replenishment occurs in the target cultivated area, and the second change value N12 is also greater than the preset water replenishment threshold T1, a verification criterion can be constructed based on the third time phase: If the second relative change value S13 is less than or equal to the second preset significant difference threshold T3, it indicates that situation A has occurred, and a precipitation event has occurred in the target cultivated area. If the second relative change value S13 is greater than the second preset significant difference threshold T3, it indicates that situation B has occurred, and an irrigation event occurred in the target cultivated area between the first and third time phases.

[0075] In summary, we have: An irrigation event occurs when C12 > T1, N12 ≤ T1, and S12 > T2; An irrigation event occurs when C12 > T1, N12 > T1, and S13 > T3; A precipitation event occurs when C12 > T1, N12 > T1, and S13 ≤ T3.

[0076] It should be noted that in other cases, it can be determined that no significant event occurred (i.e., no irrigation event or precipitation event occurred).

[0077] For example, the first preset significant difference threshold can be calibrated based on irrigation area water diversion records and precipitation records, and is set to 0.5 dB; the second preset significant difference threshold can be the same as or different from the first preset significant difference threshold, and there is no restriction on this; the second preset significant difference threshold can be adaptively determined by the first preset significant difference threshold based on the duration from the first time phase to the third time phase (see below for details); under the standard three-time phase observation window (i.e., the duration from the first time phase to the third time phase is equal to the baseline three-time phase duration), the second preset significant difference threshold is equal to the first preset significant difference threshold.

[0078] This application constructs a dual relative change value criterion for cultivated and vegetated areas by introducing VV polarization data from three consecutive time phases: utilizing the essential physical difference that irrigation events keep cultivated areas relatively moist while precipitation events cause the humidity of both areas to decrease synchronously, when the humidity of both areas rises synchronously in the dual time phase (from the first time phase to the second time phase), the second relative change value is used to verify whether the moist state of cultivated areas relative to vegetated areas continues to exist. This enables irrigation identification in scenarios that are easily missed by traditional dual time phase methods, such as irrigation with heavy precipitation and irrigation with light precipitation. Meanwhile, the irrigation event determination process in this application only uses VV polarization data and three thresholds, without relying on meteorological observation data, and without complex parameter calibration or machine learning training, which improves the efficiency of event determination and environmental adaptability (it can still be applied in areas with sparse meteorological stations or no meteorological records).

[0079] The specific differences and effects of this application compared to existing technologies are as follows: 1. When C12 > T1 and N12 > T1, the existing dual-phase method judges it as precipitation and terminates the irrigation decision, resulting in systematic missed detections in this scenario; the present invention can perform backtracking identification through the continuous verification of S13 to determine whether a rainfall event or an irrigation event has occurred.

[0080] Second, existing methods require the introduction of meteorological observation data (such as "daily precipitation > 6 mm is considered valid precipitation") as auxiliary criteria in the above two scenarios, which is limited by the distribution of meteorological stations and the timeliness of data. This invention only uses VV polarization data to complete the identification from the temporal persistence of the signal, and can still be applied in areas with sparse meteorological stations or no meteorological records.

[0081] In some possible embodiments, the SAR remote sensing identification method for farmland irrigation events resilient to precipitation interference may further include: When no significant water replenishment occurs in the target cultivated area, the fifth change value is determined based on the VV polarization data of the target cultivated area in the second and third time phases. The irrigation results for the target cultivated area are determined based on the first change value, the fifth change value, the third change value, the preset water replenishment threshold, the second relative change value, and the second preset significant difference threshold.

[0082] For example, the difference between the VV polarization data of the third and second time phases of the target cultivated area can be calculated to obtain the fifth change value C23.

[0083] It should be noted that when the VV polarization data of the target cultivated area changes in a positive direction but the magnitude is not significant (i.e., no significant water replenishment has occurred), there are two possible scenarios: Scenario C: No irrigation or effective water replenishment occurred between the first and second time phases. The slight fluctuations in VV in the cultivated area are caused by SAR spot noise or natural soil moisture changes. Scenario D: Under gradual irrigation methods such as rotational irrigation, only some fields within the target cultivated area are irrigated within a single satellite observation cycle, and the VV polarization data of the target cultivated area does not change significantly; however, after accumulating over multiple observation cycles, the changes in VV polarization data can reach a significant level.

[0084] It is understandable that scenarios C and D are indistinguishable under a single observation period (i.e., from the first phase to the second phase), but they will show essential differences in their multi-period cumulative characteristics: under scenario C, the multi-period changes are still small fluctuations with uncertain direction; under scenario D, the multi-period changes accumulate in the same direction, and the total amount reaches the threshold of the water replenishment event.

[0085] For example, a gradual irrigation event can be determined to have occurred in a target cultivated area when the following conditions are met: Condition 1: The first change value C12 is less than or equal to the preset water replenishment threshold T1 and is greater than zero (that is, the change direction of VV polarization data of the target cultivated area from the first time phase to the second time phase is positive but not significant). Condition 2: The fifth change value C23 is less than or equal to the preset water replenishment threshold T1 and is greater than zero (i.e., the change direction of VV polarization data in the target cultivated area from the second to the third time phase is positive but not significant). Condition 3: The second relative change value S13 is greater than the second preset significant difference threshold T3 (the difference between cultivated land and reference across cycles is significant, excluding the case of synchronous natural water replenishment in the reference area). Condition 4: The third change value C13 is greater than half of the preset water replenishment threshold T1 (the change in cross-cycle cumulative VV polarization data reaches the weak accumulation threshold, suppressing speckle noise misjudgment).

[0086] It should be noted that among the four conditions for identifying gradual irrigation events mentioned above: Conditions 1 and 2 ensure that both single cycles (i.e., from the first phase to the second phase, and from the second phase to the third phase) are in a state of "slow accumulation in the same direction"; Condition 3 and S13 are used as the criteria to exclude the cumulative changes from natural water replenishment caused by synchronous increases in the reference area, and to confirm the persistent wetness of the target cultivated area relative to the reference area. Condition 4 uses half of T1 as the weak accumulation threshold. This threshold is half as wide as T1, which not only preserves the physical signal characteristics of cross-cycle accumulation, but is also sufficient to suppress the false accumulation caused by the accidental unidirectional fluctuation of SAR speckle noise within two cycles.

[0087] These four conditions—single-cycle directional consistency, local specificity relative to the reference area, and cumulative magnitude significance—jointly ensure the reliability of determining gradual irrigation events.

[0088] It should be noted that condition 4 does not adopt the stringent absolute amplitude condition of "C13 > T1". This is because the cumulative signal strength of progressive irrigation (such as rotational irrigation) across cycles is affected by multiple factors such as irrigation district rotational scheduling, field coverage, and initial soil moisture. Requiring the cumulative signal strength across cycles to reach the same T1 level as single-cycle irrigation would lead to excessive missed detections of real progressive irrigation. By combining the weak cumulative threshold of condition 4 with the persistent verification of the farmland-reference difference in condition 3, a balance is achieved between suppressing noise and identifying real progressive irrigation.

[0089] This embodiment uses cross-period cumulative observation and combines the difference between cultivated land and reference as a criterion to extend the decision time window from a single period to cross periods, thus identifying gradual irrigation events that cannot be determined by the dual-temporal method.

[0090] In some possible embodiments, after determining the significant water replenishment results for the target cultivated area based on a first change value and a preset water replenishment threshold, the SAR remote sensing identification method for farmland irrigation events resistant to precipitation interference may further include: The second preset significant difference threshold is updated based on the duration of the first to third time phases and the first preset significant difference threshold.

[0091] It should be noted that as the time window (i.e., the time interval between two adjacent time phases) lengthens, the difference in VV polarization data between the target cultivated area and the reference area caused by irrigation events gradually weakens due to the following two types of physical processes: (a) Evapotranspiration attenuation: Soil moisture in the target cultivated area naturally decreases over time after irrigation, and the relative wetness of the target cultivated area compared to the reference area weakens over time. The longer the window, the weaker the relative wetness. (b) Background convergence: Natural processes such as vegetation growth, dew, and slight changes in soil moisture in the reference area have a convergent effect on the target cultivated area and the reference area over time, compressing the difference between the two.

[0092] The above mechanism shows that the longer the time window, the weaker the observable difference signal (i.e., S13) of the irrigation event, and the difference threshold (i.e., T3) can decrease monotonically with the window length to avoid missing the actual irrigation due to the natural attenuation of the signal; conversely, the shorter the time window, the tighter the threshold can be from the perspective of suppressing speckle noise interference to avoid misjudgment.

[0093] Specifically, based on the duration of the first time phase to the third time phase and the first preset significant difference threshold, the second preset significant difference threshold is updated, including: The adjustment coefficient is determined based on the duration of the first to third time phases and the duration of the reference three time phases; The second preset significant difference threshold is determined based on the adjustment coefficient and the first preset significant difference threshold.

[0094] For example, the duration of the reference three-phase system can be a preset duration determined by the satellite. For instance, the duration of the reference three-phase system for Sentinel-1 satellite can be 12 days, meaning the time interval between two adjacent phases is 6 days.

[0095] For example, the ratio of the duration of the current satellite's first to third time phases to the duration of the reference three time phases can be calculated to obtain the adjustment coefficient; then, based on the adjustment coefficient and the first preset significant difference threshold, the second preset significant difference threshold T3 can be calculated using the following formula: T3 = T2 / [1 + β × (k - 1)] In the formula, k is the adjustment coefficient; T2 is the first preset significant difference threshold; β is the time span weight coefficient, with a value range of [0,∞). When β is 0, it means no decay (i.e., T3=T2). The larger β is, the faster the threshold decay rate. In this example, the preferred value is 1.

[0096] It is understandable that when k=1 (that is, the duration from the first time phase to the third time phase is equal to the duration of the reference three time phases), T3=T2; When k=1.5 (e.g., the duration from the first to the third phase is 18 days, and the baseline three-phase duration is 12 days), T3≈0.33dB (at β=1), the threshold is relaxed to identify decayed irrigation events; When k=2 (for example, the duration from the first phase to the third phase is 24 days, and the duration of the baseline three phases is 12 days), T3≈0.25dB (at β=1), the threshold is further relaxed, but it should be noted that the threshold is close to the noise level at this time. It is recommended to combine other verification methods to ensure the reliability of identification. When k=0.5 (for example, the duration from the first phase to the third phase is 6 days, and the baseline three-phase duration is 12 days), T3=1dB (at β=1), and the threshold tightens.

[0097] This embodiment, by updating the second preset significant difference threshold, can still maintain stable event recognition capabilities even under non-uniform time intervals caused by factors such as satellite orbit adjustment and sensor collaborative observation.

[0098] In some possible embodiments, the remote sensing data also includes VH polarization data and NDVI data, and the precipitation-resistant SAR remote sensing identification method for farmland irrigation events further includes: Based on the VH polarization data of the target cultivated area and the target vegetation area in the first and second time phases, respectively, and the change values ​​of NDVI data of the target cultivated area from the first to the second time phase, the results of agricultural activities in the target cultivated area are determined.

[0099] It is understandable that VH polarization data is the backscattering coefficient data of SAR image data from remote sensing satellites after VH polarization, and it is sensitive to vegetation scattering. When agricultural activities such as harvesting and tilling occur, a large number of crop plants are removed from the farmland, and the scattering ability of vegetation is drastically reduced. This will directly manifest as a significant decrease in the backscattering coefficient of VH polarization and a decrease in the optical vegetation index (i.e., a decrease in NDVI), and the two have a strong physical correlation. In contrast, water events such as irrigation and precipitation only cause changes in SAR scattering characteristics, but NDVI remains basically stable in the short term. Through the collaborative verification of SAR and optical dual-source data, the risk of misjudgment by a single sensor can be effectively eliminated.

[0100] Traditional methods rely solely on the single feature of VH polarization data for agricultural activity identification, which is susceptible to SAR speckle noise and non-agricultural changes in vegetation structure (such as wind lodging and disease), resulting in a high risk of misjudgment and missed judgment. This embodiment proposes a dual-modal discrimination mechanism that combines VH polarization and NDVI. Agricultural activity is only identified when both SAR features (significant decrease in VH) and optical features (significant decrease in NDVI) are met simultaneously, which helps reduce the uncertainty of single-modal discrimination.

[0101] For example, the difference between the VH polarization data of the target cultivated area in the second time phase and the first time phase can be calculated to obtain the sixth change value VH_C12; the difference between the VH polarization data of the target vegetation area in the second time phase and the first time phase can be calculated to obtain the seventh change value VH_N12; the difference between the NDVI data of the target cultivated area in the second time phase and the first time phase can be calculated to obtain the eighth change value NDVI_C12; and the difference between the sixth change value and the seventh change value can be calculated to obtain the third relative change value VH_S12. When the sixth change value VH_C12 is less than the preset vegetation removal threshold T4 (indicating a significant decrease in VH polarization data in the target cultivated area), the third relative change value VH_S12 is less than the third preset significant difference threshold T5 (indicating a significantly greater decrease in VH polarization data in the target cultivated area than in the reference area), and the eighth change value NDVI_C12 is less than the preset vegetation coverage threshold T6 (indicating a significant decrease in NDVI data in the target cultivated area; NDVI data is the most commonly used vegetation coverage indicator in optical remote sensing. After crop harvesting, farmland vegetation coverage will drop sharply from a high value to a low value, which is reflected in a significant decrease in NDVI data; this optical independent verification condition solves the problem of misjudgment of single SAR data), it can be determined that agricultural activities occurred in the target cultivated area from the first time phase to the second time phase.

[0102] For example, the preset vegetation removal threshold T4 can be -1.5 dB, the third preset significant difference threshold T5 can be -1.0 dB, and the preset vegetation coverage threshold T6 can be -0.1.

[0103] It should be noted that if an irrigation event is confirmed to have occurred in the target cultivated area in the aforementioned steps, and agricultural activities are confirmed to have taken place in the target cultivated area, then it is ultimately confirmed that no irrigation event has occurred in the target cultivated area.

[0104] This embodiment combines VH polarization data and NDVI data to determine the results of agricultural activities in the target cultivated area, which can then be used to correct irrigation results and improve the accuracy of determining irrigation results in the target cultivated area.

[0105] The technical advantages of this application are further illustrated below through a specific embodiment.

[0106] For a target irrigation area in Puyang City, Henan Province, the method described in this application was used to identify irrigation events from the Sentinel-1 SAR time series (including VV polarization data and VH polarization data) during the winter wheat growing season (January 1 to June 6, 2025). All three-phase observations were performed in the standard single-satellite operating mode of Sentinel-1, with an interval of 6 days between adjacent phases and an overall span of 12 days from the first to the third phase (k=1), with T3=T2=0.5 dB.

[0107] Overall identification results: Five precipitation events (January 19-January 26, February 12-February 19, February 24-March 3, April 26-May 2, May 7-May 14) and four irrigation events (February 12-February 24, April 14-April 25, April 26-May 7, May 7-May 19) were correctly identified, and the identified irrigation periods were consistent with the irrigation district's water diversion records.

[0108] Identification of precipitation-irrigation complex events: The target irrigation district experienced a combined event of precipitation and irrigation occurring simultaneously between April 26 and May 7, 2025.

[0109] Using the traditional two-phase method: From April 26 (phase 1) to May 2 (phase 2), the VV polarization data of both the cultivated area and the reference area increased significantly (C12 > T1 and N12 > T1). The traditional method would identify this as a precipitation event and terminate the irrigation decision, resulting in missed irrigation detection. However, if meteorological data (cumulative precipitation from phase 1 to phase 2 reaching 6 mm) is further introduced, the "effective precipitation" determination would be further confirmed, thus ruling out the possibility of irrigation. Using the method of this application: under the condition that C12>T1 and N12>T1, May 7th is introduced as the third time phase, and S13=C13-N13>0.5 dB is calculated to determine that an irrigation event occurred between April 26th and May 7th, which is consistent with the irrigation district water diversion record.

[0110] Under the scenario of two light rainfalls superimposed on irrigation: During the period from February 12 to February 19, light precipitation caused N12 > T1 and S13 > T3 to remain significantly positive, which was identified as an irrigation event; during the period from May 7 to May 14, similar processes determined that no effective precipitation event equivalent to irrigation occurred during this period.

[0111] The three composite events mentioned above would all be missed by the traditional two-phase method, but this method correctly identified them all by continuous verification in the third phase.

[0112] Cumulative identification of progressive irrigation: A gradual rotational irrigation event occurred in the target irrigation district from May 7 to May 19, 2025, in which fields within the area were irrigated in batches in sequence.

[0113] Using the traditional two-phase method: During the period from May 7 (first phase) to May 13 (second phase), there is a period of 0 < C12 ≤ T1 (only some fields have been irrigated within a single cycle). The traditional method judges this as no event, resulting in missed detections. Using the method described in this application: Under the condition 0 < C12 ≤ T1, May 19th is introduced as the third time phase to test four conditions: (i) C12=0.56 dB∈(0,T1]; (ii) C23=0.42 dB∈(0,T1]; (iii) C13=0.98 dB>T1 / 2=0.5 dB; (iv) S13=1.03 dB>T3=0.5 dB; It can be seen that all four conditions are met, confirming that a gradual irrigation event occurred between May 7th and May 19th, which is consistent with the irrigation district's rotational irrigation records.

[0114] It is understandable that after obtaining the final irrigation results, the irrigation event identification results of each time phase can be summarized, and the output includes: (1) Spatiotemporal distribution of irrigation events: Spatial distribution map of irrigation areas identified in each time phase; (2) Irrigation frequency grid product: Statistics on the number of irrigations per pixel / plot during the monitoring period; (3) Actual irrigated area statistics: The total area of ​​cultivated land that has been irrigated at least once during the monitoring period.

[0115] The irrigation results for the target irrigation district during the winter wheat growing season in 2025 were as follows: the actual irrigated area was 84,000 mu (96.6% agreement with the effective irrigated area of ​​87,000 mu); the irrigation frequency grid product showed an average of 6.5 irrigations (this value is a statistical measure at the pixel / plot scale, reflecting the cumulative number of identifications at different pixel phases due to factors such as rotational irrigation and staggered irrigation within the same irrigation district), with 69.8% of the cultivated land receiving 5-7 irrigations; spatially, the northern region averaged 6.6 irrigations, and the southern region averaged 6.2 irrigations. The resulting grid distribution map of irrigation frequency during the growing season is shown below. Figure 3 As shown.

[0116] This application also provides an application scenario in which the above-mentioned SAR remote sensing identification method for farmland irrigation events with resistance to precipitation interference is applied. Specifically, the SAR remote sensing identification method for farmland irrigation events with resistance to precipitation interference provided in this embodiment can be applied in agricultural production guidance scenarios. First, the irrigation frequency of the cultivated area is obtained, then crop growth and yield are predicted based on the irrigation frequency of the cultivated area, and differentiated fertilization plans are customized according to the irrigation frequency of different cultivated areas. The SAR remote sensing identification method for farmland irrigation events with resistance to precipitation interference provided in this embodiment belongs to the stage of obtaining the irrigation frequency of cultivated areas.

[0117] Based on the same inventive concept, this application also provides a precipitation-resistant SAR remote sensing identification system for implementing the precipitation-resistant SAR remote sensing identification method for farmland irrigation events described above. The solution provided by this system is similar to the solution described in the above method. Therefore, the specific limitations of one or more precipitation-resistant SAR remote sensing identification system embodiments provided below can be found in the limitations of the precipitation-resistant SAR remote sensing identification method described above, and will not be repeated here.

[0118] In one exemplary embodiment, such as Figure 4 As shown, a SAR remote sensing identification system for farmland irrigation events with resistance to precipitation interference is provided, including: The acquisition module 401 is used to acquire remote sensing data of multiple pixels in the target irrigation area and the land type corresponding to the multiple pixels. The remote sensing data includes VV polarization data, and the land type includes cultivated area and vegetation area. The calculation module 402 is used to determine the first change value, the second change value, the third change value, the fourth change value, the first relative change value, and the second relative change value based on the VV polarization data of the target cultivated area and the target vegetation area in the first, second, and third consecutive time phases, respectively. The first change value is the change in VV polarization data of the target cultivated area from the first time phase to the second time phase; the second change value is the change in VV polarization data of the target vegetation area from the first time phase to the second time phase; the third change value is the change in VV polarization data of the target cultivated area from the first time phase to the third time phase; and the fourth change value is the change in VV polarization data of the target vegetation area from the first time phase to the third time phase. The first relative change value is the difference between the first change value and the second change value; and the second relative change value is the difference between the third change value and the fourth change value. The determination module 403 is used to determine the significant water replenishment result of the target cultivated area based on the first change value and the preset water replenishment threshold; when significant water replenishment occurs in the target cultivated area, the irrigation result of the target cultivated area is determined based on the second change value, the preset water replenishment threshold, the first relative change value, the first preset significant difference threshold, the second relative change value, and the second preset significant difference threshold. Determining the irrigation results of the target cultivated area includes: when the second change value is greater than the preset water replenishment threshold, determining whether an irrigation event has occurred in the target cultivated area from the first time phase to the third time phase based on the second relative change value and the second preset significant difference threshold.

[0119] The specific implementation methods and beneficial effects of this system embodiment can be found in the foregoing method embodiments, and will not be repeated here.

[0120] In one exemplary embodiment, a computer device is provided, which may be a server or a terminal, and its internal structure diagram may be as follows. Figure 5 As shown, the computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores data. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements the aforementioned SAR remote sensing identification method for farmland irrigation events resistant to precipitation interference.

[0121] Those skilled in the art will understand that Figure 5 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0122] In one exemplary embodiment, a computer device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0123] In one exemplary embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0124] In one exemplary embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0125] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0126] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this 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), magnetic 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. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0127] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0128] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0129] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A SAR remote sensing method for identifying farmland irrigation events resistant to precipitation interference, characterized in that, The SAR remote sensing identification method for farmland irrigation events that is resistant to precipitation interference includes: Acquire remote sensing data of multiple pixels in the target irrigation area and the land types corresponding to the multiple pixels. The remote sensing data includes VV polarization data, and the land types include cultivated areas and vegetated areas. Based on the VV polarization data of the target cultivated area and the target vegetation area in the first, second, and third consecutive time phases, the first change value, the second change value, the third change value, the fourth change value, the first relative change value, and the second relative change value are determined. Wherein, the first change value is the change value of VV polarization data of the target cultivated area from the first time phase to the second time phase; the second change value is the change value of VV polarization data of the target vegetation area from the first time phase to the second time phase; the third change value is the change value of VV polarization data of the target cultivated area from the first time phase to the third time phase; and the fourth change value is the change value of VV polarization data of the target vegetation area from the first time phase to the third time phase. The first relative change value is the difference between the first change value and the second change value; and the second relative change value is the difference between the third change value and the fourth change value. Based on the first change value and the preset water replenishment threshold, the significant water replenishment results for the target cultivated area are determined; When significant water replenishment occurs in the target cultivated area, the irrigation result of the target cultivated area is determined based on the second change value, the preset water replenishment threshold, the first relative change value, the first preset significant difference threshold, the second relative change value, and the second preset significant difference threshold. Determining the irrigation results of the target cultivated area includes: when the second change value is greater than the preset water replenishment threshold, determining whether an irrigation event has occurred in the target cultivated area from the first time phase to the third time phase based on the second relative change value and the second preset significant difference threshold.

2. The SAR remote sensing identification method for farmland irrigation events with resistance to precipitation interference according to claim 1, characterized in that, The determination of the land type corresponding to the multiple pixels includes: Based on the historical NDVI data of the multiple pixels, determine the maximum NDVI data of the multiple pixels during the first preset period before the crop harvest period, and the minimum NDVI data of the multiple pixels during the crop harvest period; The harvest index is determined based on the maximum NDVI data and the minimum NDVI data; Based on the harvest index, determine the first division threshold and the second division threshold; Based on the harvest index, the first classification threshold, and the second classification threshold, the land type corresponding to the plurality of pixels is determined.

3. The SAR remote sensing identification method for farmland irrigation events with resistance to precipitation interference according to claim 1, characterized in that, The SAR remote sensing identification method for farmland irrigation events with resistance to precipitation interference also includes: When no significant water replenishment occurs in the target cultivated area, the fifth change value is determined based on the VV polarization data of the target cultivated area in the second and third time phases. The irrigation results for the target cultivated area are determined based on the first change value, the fifth change value, the third change value, the preset water replenishment threshold, the second relative change value, and the second preset significant difference threshold.

4. The SAR remote sensing identification method for farmland irrigation events with resistance to precipitation interference according to claim 1, characterized in that, After determining the significant water replenishment results for the target cultivated area based on the first change value and the preset water replenishment threshold, the SAR remote sensing identification method for farmland irrigation events resistant to precipitation interference further includes: The second preset significant difference threshold is updated based on the duration of the first to third time phases and the first preset significant difference threshold.

5. The SAR remote sensing identification method for farmland irrigation events with resistance to precipitation interference according to claim 4, characterized in that, The step of updating the second preset significant difference threshold based on the duration from the first time phase to the third time phase and the first preset significant difference threshold includes: The adjustment coefficient is determined based on the duration of the first to third time phases and the duration of the reference three time phases; The second preset significant difference threshold is determined based on the adjustment coefficient and the first preset significant difference threshold.

6. The SAR remote sensing identification method for farmland irrigation events with resistance to precipitation interference according to claim 1, characterized in that, The remote sensing data also includes VH polarization data and NDVI data, and the SAR remote sensing identification method for farmland irrigation events resistant to precipitation interference further includes: Based on the VH polarization data of the target cultivated area and the target vegetation area in the first and second time phases, respectively, and the change values ​​of NDVI data of the target cultivated area from the first to the second time phase, the results of agricultural activities in the target cultivated area are determined.

7. A SAR remote sensing identification system for farmland irrigation events resistant to precipitation interference, characterized in that, The SAR remote sensing identification system for farmland irrigation events that is resistant to precipitation interference includes: The acquisition module is used to acquire remote sensing data of multiple pixels in the target irrigation area and the land type corresponding to the multiple pixels. The remote sensing data includes VV polarization data, and the land type includes cultivated area and vegetation area. The calculation module is used to determine the first change value, the second change value, the third change value, the fourth change value, the first relative change value, and the second relative change value based on the VV polarization data of the target cultivated area and the target vegetation area in the first, second, and third consecutive time phases, respectively. Wherein, the first change value is the change value of VV polarization data of the target cultivated area from the first time phase to the second time phase; the second change value is the change value of VV polarization data of the target vegetation area from the first time phase to the second time phase; the third change value is the change value of VV polarization data of the target cultivated area from the first time phase to the third time phase; and the fourth change value is the change value of VV polarization data of the target vegetation area from the first time phase to the third time phase. The first relative change value is the difference between the first change value and the second change value; and the second relative change value is the difference between the third change value and the fourth change value. The determination module is used to determine the significant water replenishment result of the target cultivated area based on the first change value and the preset water replenishment threshold; when significant water replenishment occurs in the target cultivated area, the irrigation result of the target cultivated area is determined based on the second change value, the preset water replenishment threshold, the first relative change value, the first preset significant difference threshold, the second relative change value, and the second preset significant difference threshold. Determining the irrigation results of the target cultivated area includes: when the second change value is greater than the preset water replenishment threshold, determining whether an irrigation event has occurred in the target cultivated area from the first time phase to the third time phase based on the second relative change value and the second preset significant difference threshold.

8. A computer device, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the computer program to implement the steps of the SAR remote sensing identification method for farmland irrigation events resistant to precipitation interference as described in any one of claims 1-6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the steps of the SAR remote sensing identification method for farmland irrigation events with resistance to precipitation interference as described in any one of claims 1-6.

10. A computer program product, comprising a computer program, characterized in that, When executed by a processor, the computer program implements the steps of the SAR remote sensing identification method for farmland irrigation events with resistance to precipitation interference as described in any one of claims 1-6.

Images