Method and device for agricultural insurance and nuclear insurance, electronic equipment and storage medium

Abstract

This invention relates to the field of artificial intelligence technology and discloses an agricultural insurance underwriting method, apparatus, electronic device, and storage medium. The method includes: receiving regional location information, agricultural category, and underwriting mode of insured objects in a target insured area; collecting meteorological disaster data, remote sensing image data, and historical agricultural insurance claim data based on the regional location information and agricultural category; determining the current risk level of the target insured area based on the meteorological disaster data, remote sensing image data, and historical agricultural insurance claim data; selecting a chosen sampling rule from candidate sampling rules based on the current risk level, regional location information, and agricultural category; sending the chosen sampling rule to an agricultural insurance sampling terminal, and receiving target image data of the insured objects from the agricultural insurance sampling terminal based on the chosen sampling rule; and performing underwriting for the target insured area based on the underwriting mode and target image data. This method can be applied to agricultural insurance scenarios in fintech, saving underwriting manpower and improving underwriting accuracy and efficiency.

Description

Technical Field

[0001] This invention relates to the fields of artificial intelligence technology and natural language processing technology, and is applied to financial technology scenarios. In particular, it relates to an agricultural insurance underwriting method and device, electronic device and storage medium. Background Technology

[0002] Agricultural insurance underwriting primarily involves post-disaster loss assessment, requiring the selection of representative insured agricultural plots for yield measurement to estimate the overall loss across the entire region and complete the underwriting process. In related technologies, the verification of insured agricultural plots mainly employs an offline, manual underwriting model, where local personnel responsible for the insured area complete the underwriting process with paper signatures and manual photographs. Furthermore, the sampling strategy for agricultural insurance underwriting relies on fixed sampling ratios or manually designated sampling targets, resulting in insufficient sample representativeness and an inability to objectively reflect the actual condition of agricultural plots in the region. Simultaneously, manually setting sampling ratios fails to differentiate sampling based on plot risk, treating low-risk and high-risk plots the same, leading to an unreasonable allocation of underwriting resources. The inventors recognized that the aforementioned manual agricultural insurance underwriting methods suffer from two shortcomings: firstly, insufficient sample representativeness, making it impossible to set sampling ratios or sampling targets based on plot risk differences; and secondly, the high manpower consumption associated with manual agricultural insurance underwriting. Summary of the Invention

[0003] This invention provides a method and apparatus for agricultural insurance underwriting, an electronic device and a storage medium to solve the technical problems of high manpower consumption, low accuracy and low efficiency in the agricultural insurance underwriting process.

[0004] Firstly, a method for agricultural insurance underwriting is provided, including: Receive agricultural insurance underwriting requests for the target insured area; wherein, the target insured area is the area where the insured is located, and the agricultural insurance underwriting request includes: the insured's regional location information, agricultural category, and insurance mode; Based on the regional location information and the agricultural category, meteorological disaster data, remote sensing image data, and historical agricultural insurance compensation data of the target insured area are collected. A risk assessment is conducted on the target insured area based on the meteorological disaster data, the remote sensing image data, and the historical agricultural insurance compensation data to obtain the current risk level; Based on the current risk level, the regional location information, and the agricultural category, a selected sampling rule is selected from the preset candidate sampling rules; wherein, the selected sampling rule represents the sampling ratio and image sampling requirements of the target insured area; The selected sampling rule is sent to the agricultural insurance sampling terminal, and the target image data of the insured object is received from the agricultural insurance sampling terminal according to the selected sampling rule. The target insured area is underwritten based on the underwriting model and the target image data.

[0005] Secondly, an agricultural insurance underwriting device is provided, comprising: The request receiving module is used to receive agricultural insurance underwriting requests for the target insured area; wherein, the target insured area is the area where the insured is located, and the agricultural insurance underwriting request includes: the insured's regional location information, agricultural category, and insurance mode; The data acquisition module is used to collect meteorological disaster data, remote sensing image data and historical agricultural insurance compensation data of the target insured area based on the regional location information and the agricultural category; The risk assessment module is used to conduct a risk assessment of the target insured area based on the meteorological disaster data, the remote sensing image data, and the historical agricultural insurance compensation data, and to obtain the current risk level. The rule filtering module is used to filter out selected sampling rules from preset candidate sampling rules based on the current risk level, the regional location information, and the agricultural category; wherein, the selected sampling rule represents the sampling ratio and image sampling requirements of the target insured area; The transceiver module is used to send the selected sampling rule to the agricultural insurance sampling terminal and receive the target image data of the insured object fed back by the agricultural insurance sampling terminal according to the selected sampling rule; The underwriting module is used to underwrite the target insured area based on the underwriting mode and the target image data.

[0006] Thirdly, a computer device is provided, 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 above-mentioned agricultural insurance underwriting method.

[0007] Fourthly, a computer-readable storage medium is provided, which stores a computer program that, when executed by a processor, implements the steps of the above-mentioned agricultural insurance underwriting method.

[0008] The aforementioned agricultural insurance underwriting method, device, electronic equipment, and storage medium can receive underwriting requests containing regional location information, agricultural category, and underwriting mode. It combines three types of multi-source heterogeneous data—meteorological disaster data, remote sensing imagery, and historical claims data—and adaptively selects sampling rules matching regional characteristics and agricultural type based on the dynamic risk level obtained from the fusion assessment. These rules are then precisely distributed to the sampling end, which receives the target image data in return. Finally, the underwriting determination is completed by combining the underwriting mode. Therefore, this embodiment achieves scientific quantification of land risk through multi-source data fusion assessment, saves underwriting manpower, and achieves differentiated and precise sampling through risk-driven dynamic sampling rules, improving underwriting efficiency and accuracy. Attached Figure Description

[0009] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0010] Figure 1 This is a schematic diagram of an application environment for the agricultural insurance underwriting method in one embodiment of the present invention; Figure 2 This is a flowchart illustrating the agricultural insurance underwriting method in one embodiment of the present invention; Figure 3 This is a schematic diagram of a scenario of the agricultural insurance underwriting method in one embodiment of the present invention; Figure 4 yes Figure 2 A schematic diagram of a specific implementation method for step S30; Figure 5 yes Figure 4 A flowchart illustrating a specific implementation of step S34; Figure 6 yes Figure 2 A schematic diagram of a specific implementation of step S40; Figure 7 yes Figure 2 A schematic diagram of a specific implementation method for step S50; Figure 8 yes Figure 7 A schematic diagram of a specific implementation method for step S52; Figure 9 This is a flowchart illustrating a method for agricultural insurance underwriting in another embodiment of the present invention; Figure 10 This is a schematic diagram of the agricultural insurance underwriting device in one embodiment of the present invention; Figure 11This is a schematic diagram of the structure of a computer device according to an embodiment of the present invention; Figure 12 This is another structural schematic diagram of a computer device according to one embodiment of the present invention. Detailed Implementation

[0011] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0012] The agricultural insurance underwriting method provided in this embodiment of the invention can be applied to, for example... Figure 1 In the application environment, the system includes an underwriting terminal 101, a server terminal 103, and at least one agricultural insurance sampling terminal 104. The underwriting terminal 101 communicates with the server terminal 103 via network 102, and the agricultural insurance sampling terminal 104 also communicates with the server terminal 103 via network 102. The underwriting terminal 101 is used to upload agricultural insurance underwriting requests and is the terminal used by the underwriter to complete insurance verification when the claims process is initiated. If the insured suffers losses due to natural disasters or other reasons, the policyholder sends a claims request to the underwriting terminal 101, and the underwriter outputs the agricultural insurance underwriting request to the server terminal 103. The server terminal 103 is equipped with an agricultural insurance underwriting platform, which retrieves the target image data of the insured in the target insured area based on the agricultural insurance underwriting request. It should be noted that, in order to reduce manpower for sampling and improve underwriting accuracy, the risk level of the target insured area is initially assessed. Then, based on the risk level, the location information of the target insured area, and the agricultural category, sampling ratios and image sampling parameters are set specifically, and target image data of the insured objects is collected in a targeted manner. This improves the efficiency and accuracy of agricultural insurance underwriting and reduces excessive manpower consumption. It should also be noted that the target image data is generated by the agricultural insurance underwriting platform by combining the sampling ratio and image sampling parameters to form sampling rules, which are then sent to the agricultural insurance sampling terminal 104. The agricultural insurance sampling terminal 104 collects target image data in a targeted manner according to the sampling rules, improving the utilization and effectiveness of target image data in the agricultural insurance underwriting process, improving the accuracy and efficiency of agricultural insurance underwriting, and saving manpower in agricultural insurance underwriting. The underwriting terminal 101 and the agricultural insurance sampling terminal 104 can be, but are not limited to, various personal computers, laptops, smartphones, tablets, and portable wearable devices. The server terminal 103 can be implemented using a separate server or a server cluster consisting of multiple servers. The invention will be described in detail below through specific embodiments.

[0013] Please see Figure 2 As shown, Figure 2A flowchart illustrating the agricultural insurance underwriting method provided in this embodiment of the invention includes the following steps: S10: Receive agricultural insurance underwriting requests for the target insured area; where the target insured area is the area where the insured is located, and the agricultural insurance underwriting request includes: the insured's regional location information, agricultural category and insurance mode.

[0014] The agricultural insurance underwriting method provided by this invention is applied in an agricultural insurance underwriting engine. The agricultural insurance underwriting engine communicates with the underwriting terminal to receive agricultural insurance underwriting requests. The agricultural insurance underwriting request includes the regional location information of the target insured area, the agricultural category of the insured, and the underwriting mode. It should be noted that agricultural insurance is mainly divided into crop insurance and livestock insurance. Agricultural categories include crop categories and livestock categories. Reference objects include crops, flowers, trees, medicinal herbs, livestock, poultry, and aquatic products. This embodiment does not limit the reference objects. The underwriting mode represents the underwriting operation of the insured, specifically including individual underwriting mode, collective underwriting mode, and individual underwriting mode. Individual underwriting is a one-to-one underwriting mode for large-scale operating entities. The collective underwriting mode is for ordinary farmers with scattered planting or breeding operations, with unified underwriting at the village or township level. The individual underwriting mode is the mode where individual farmers purchase insurance on an insurance business platform.

[0015] Furthermore, after an insured person initiates a claim, the underwriter initiates an agricultural insurance underwriting request for the target insured area. This request is used to verify the insured individuals within the target area, serving as a pre-claim information verification process. Figure 3 As shown, Town A is a major lychee producer. Agricultural insurance has been purchased for all lychee farmers in the town. However, this year, due to rainfall and weather conditions, the lychee yield in Town A is low. The agricultural insurance policy covers lychee yields below a predetermined value. Farmers in Town A have filed claims. Since claims require verification of Town A's lychee production and this year's planting conditions, underwriting is necessary in advance as reference data. Therefore, claims personnel initiated an agricultural insurance underwriting request to the agricultural insurance underwriting engine, which first analyzed the risk situation in Town A and then generated a targeted sampling mechanism for Town A.

[0016] S20: Collect meteorological disaster data, remote sensing image data, and historical agricultural insurance compensation data for the target insured area based on regional location information and agricultural category.

[0017] It should be noted that, in order to assess the risk of the target insured area, the risk of the insured in the target insured area will be comprehensively analyzed from three aspects: climate disasters, current remote sensing images, and historical claims data, thereby improving the accuracy of risk assessment. Specifically, the agricultural insurance underwriting engine obtains climate disaster data of the target insured area during the underwriting period from the climate disaster website based on regional location information and agricultural category, and then collects remote sensing image data from remote sensing equipment based on regional location information and agricultural category. The remote sensing equipment includes satellites, drones, remote monitoring, and remote thermal imaging sensors, etc., and this embodiment does not limit the remote sensing equipment. Furthermore, based on regional location information and agricultural category, historical agricultural insurance claims data of the target insured area is obtained from the historical claims database of the insurance business platform.

[0018] S30: Conduct a risk assessment of the target insured area based on meteorological disaster data, remote sensing image data, and historical agricultural insurance compensation data to obtain the current risk level.

[0019] It is important to understand that the risk assessment here refers to the payout risk of insured individuals within the target insured area, enabling advance prediction of claims and rapid underwriting when claims are needed, thus improving claims efficiency. Specifically, this embodiment breaks away from the rigid traditional model of "fixed proportions and manual designation," integrating multi-dimensional information such as historical agricultural insurance payout data, meteorological disaster data, remote sensing imagery data, and historical agricultural insurance payout data to generate risk profiles of farmers and the target insured area, which are then used as the basis for sampling decisions. In some embodiments, please refer to... Figure 4 Step S30 may include the following steps: S31: Based on meteorological disaster data, risk prediction is performed on the target insured area to obtain the first risk value.

[0020] S32: Based on remote sensing image data, risk prediction is performed on the target insured area to obtain a second risk value.

[0021] S33: Based on historical agricultural insurance compensation data, risk prediction is performed on the target insured area to obtain the third risk value.

[0022] S34: Conduct a risk assessment of the target insured area based on the first risk value, the second risk value, and the third risk value to obtain the current risk level.

[0023] For steps S31-S33, climate disaster data includes historical rainfall, temperature, and extreme weather events. If the insured object is crops, remote sensing image data represents spatial information such as crop growth and plot boundaries; if the insured object is livestock, remote sensing image data represents livestock growth trends and breeding environment. Historical agricultural insurance claim data records the frequency of claims and the amount of compensation in the target insured area, directly reflecting the historical risk status of the target insured area.

[0024] Climate disaster characteristics are extracted from climate disaster data. A climate disaster index is obtained by predicting climate disasters in the target insured area using a disaster prediction model and these characteristics. The first risk value is determined based on this climate disaster index. Remote sensing image characteristics of the target insured area are extracted from remote sensing image data. A remote sensing growth index is obtained by predicting the growth of insured individuals using a remote sensing growth prediction model and these image characteristics. The second risk value is determined based on this remote sensing growth index. Historical payout rates are extracted from historical agricultural insurance payout data. The third risk value is determined based on these historical payout rates.

[0025] In step S34, the first risk value, the second risk value, and the third risk value are concatenated to determine the current risk value of the target insured area, and then the current risk level is determined based on the current risk value. Specifically, in this embodiment, the concatenation of the first risk value, the second risk value, and the third risk value is a weighted concatenation, constructing a more accurate risk level representing the risk of the insured persons in the target insured area. For details, please refer to... Figure 5 Step S34, which involves assessing the risk of the target insured area based on the first risk value, the second risk value, and the third risk value to obtain the current risk level, includes the following steps: S341: Assign a first weight to the first risk value, a second weight to the second risk value, and a third weight to the third risk value.

[0026] S342: The first risk value, the second risk value, and the third risk value are concatenated according to the first weight, the second weight, and the third weight to obtain the current risk value.

[0027] S343: Determine the current risk level based on the current risk value and the preset risk threshold.

[0028] For step S341, the first weight, second weight, and third weight can be set based on expert weight allocation experience, or based on historical weight allocation information. This embodiment can also combine a hierarchical model to perform importance analysis on meteorological disaster data, remote sensing image data, and historical agricultural insurance compensation data to obtain importance analysis data, and set the first weight, second weight, and third weight based on this importance analysis data. In this regard, this embodiment does not limit the allocation rules for the first weight, second weight, and third weight.

[0029] For step S342, define the first weight as Q1, the second weight as Q2, the third weight as Q3, the first score as F1, the second score as F2, the third score as F3, and determine the current risk value as Fz = Q1×F1 + Q2×F2 + Q2×F3.

[0030] For step S343, a risk threshold is set for each risk level, and the risk level corresponding to the current risk value being greater than the risk threshold is taken as the current risk level. If this embodiment sets three risk levels: low risk, medium risk, and high risk, the risk threshold between the low and medium risk levels is the first risk threshold, and the risk threshold between the medium and high risk levels is the second risk threshold. The first risk threshold is less than the second risk threshold. If the current risk value is greater than the first risk threshold, the current risk level is determined to be medium risk; if the current risk value is less than the first risk threshold, the current risk level is determined to be low risk; if the current risk value is greater than the second risk threshold, the current risk level is determined to be high risk.

[0031] The above scheme assigns independent weights to risk values ​​from different sources and merges them using a splicing process rather than simple summation. This approach preserves the independence of risk characteristics across different dimensions while achieving a precise and objective quantification of the overall risk status of a land parcel. This dynamic weighting and feature splicing mechanism effectively overcomes the technical shortcomings of traditional manual experience-based assessments, which are characterized by strong subjectivity and limited dimensions, significantly improving the scientific rigor and accuracy of risk level classification.

[0032] Furthermore, through steps S31-S34, based on meteorological disaster data, remote sensing imagery data, and historical agricultural insurance compensation data, the first, second, and third risk values ​​are predicted, respectively. A comprehensive risk assessment is then conducted based on these values, constructing a multi-dimensional and three-dimensional risk perception system. This mechanism of parallel processing and fusion assessment of multi-source data effectively breaks through the limitations of traditional single-source data source risk assessment. Meteorological disaster data can reflect the region's climate stress risk in real time; remote sensing imagery data can objectively characterize the actual growth status and spatial distribution heterogeneity of crops; and historical compensation data implicitly contains long-term accumulated risk patterns and land vulnerability information. The organic combination and cross-validation of these three elements ensure that the final determined current risk level possesses both the ability to quickly respond to instantaneous meteorological shocks and integrates spatial differences in crop growth information and statistical patterns of historical risks, thereby significantly improving the scientific rigor, comprehensiveness, and accuracy of the risk assessment.

[0033] S40: Select a sampling rule from the preset candidate sampling rules based on the current risk level, regional location information, and agricultural category; wherein, the selected sampling rule represents the sampling ratio and image sampling requirements of the target insured area.

[0034] It should be noted that traditional sampling relies on fixed proportions or manual designation, lacking the ability to dynamically adjust for regional risks, historical data, and spatiotemporal characteristics. This results in insufficient representativeness of the sampled data, failing to reflect the true situation of the target insured area. Furthermore, the sampling strategy lacks tree algorithm support and does not incorporate mechanisms for randomness, coverage, and bias control, leading to distorted sample data and low accuracy in agricultural insurance underwriting. To address this, this embodiment combines the current risk level, regional location information, and agricultural insurance category as the basis for sampling rules. The agricultural insurance underwriting engine dynamically adjusts the sampling proportion and image acquisition requirements, transforming candidate sampling rules into selected ones. It should be noted that the sampling rules in this embodiment are set according to the principle of "more sampling for high-risk areas, fewer sampling for low-risk areas, and balanced regional coverage." Therefore, this embodiment extracts images of insured individuals in the target insured area according to the selected sampling rules, providing a more accurate representation of the insured individuals' situation.

[0035] In some embodiments, please refer to Figure 6 Step S40, which involves selecting a sampling rule from the preset candidate sampling rules based on the current risk level, regional location information, and agricultural category, includes the following steps: S41: Select the chosen sampling ratio from the preset candidate sampling ratios based on the current risk level.

[0036] S42: Select the image sampling parameters from the preset candidate image sampling parameters based on the regional location information and agricultural category.

[0037] S43: Concatenate the selected sampling ratio and the selected image sampling parameters to obtain the selected sampling rule.

[0038] In step S41, a first mapping relationship table between each risk level and the candidate sampling ratio is set in advance. The candidate sampling ratio corresponding to the current risk level is determined as the selected sampling ratio based on the first mapping relationship table. For example, if the first mapping relationship records that the candidate sampling ratio for high-risk level is 70%, the candidate sampling ratio for medium-risk level is 30%, and the candidate sampling ratio for low-risk level is 5%.

[0039] In step S42, a second mapping table is pre-set for the location information of each region, agricultural category, and candidate image sampling parameters. The candidate image sampling parameters corresponding to the region location information and agricultural category are then retrieved from the second mapping table and used as the selected image sampling parameters. These selected image sampling parameters include setting location information, farmer information, sampling angle information, and sampling clarity information. This embodiment does not limit the selected image sampling parameters; their content can be set according to requirements. Furthermore, the selected image sampling parameters serve as guidance information for taking images of insured individuals.

[0040] In step S43, the selected sampling ratio and the selected image sampling parameters are concatenated into a selected sampling rule, which serves as an indication for subsequent target image data sampling.

[0041] For example, if policyholders in Town A have good credit, the current risk level for lychees in Town A is assessed as medium risk; if the historical claim rate for lychees in Town B is high, the current risk level for lychees in Town B is assessed as high risk. Therefore, it is determined that 50% of the lychee image data from Town A will be collected as the target image data, while only 70% of the lychee image data from Town B will be collected as the target image data. Thus, a larger proportion of samples needs to be collected for high-risk areas to improve underwriting accuracy, while only a portion of the samples is needed for low-risk areas to achieve accurate underwriting and improve underwriting targeting.

[0042] The above scheme, through the organic combination of risk-driven dynamic proportion control and spatially-aware adaptive parameter selection, significantly improves the representativeness, pertinence and flexibility of agricultural insurance verification sampling, effectively enhances the authenticity and reliability of sampling results, provides a high-quality data foundation for subsequent compliance verification, land ownership confirmation and accurate claims settlement based on sampling data, and further consolidates the management closed-loop capability of the smart agricultural insurance platform.

[0043] S50: Send the selected sampling rule to the agricultural insurance sampling terminal and receive the target image data of the insured object from the agricultural insurance sampling terminal according to the selected sampling rule.

[0044] In this embodiment, at least two candidate sampling terminals are set up. It is also necessary to collect the work area information, current location information, and work load information of each candidate sampling terminal. A first score is assigned to the agricultural insurance sampling terminal based on the work area information, current location information, agricultural category, and regional location information. A second score is assigned based on the work load information. The first score and the second score are weighted and averaged to obtain the adaptation score of each agricultural insurance sampling terminal. The adaptation score represents the sampling adaptation degree of the candidate sampling terminal to the target insured area. Then, the agricultural insurance sampling terminal is selected from the candidate sampling terminals based on the adaptation score, and the selected sampling rule is sent to the agricultural insurance sampling terminal.

[0045] Furthermore, traditional sampling methods do not verify the identity and behavior of sampling personnel at the agricultural insurance sampling terminal, leading to doubts about the authenticity of uploaded sampling image data and the possibility of forged image data of insured individuals. Therefore, this embodiment requires verifying the agricultural insurance sampling terminal before receiving target image data uploaded by verified sampling terminals.

[0046] In some embodiments, please refer to Figure 7 Step S50, which involves sending the selected sampling rule to the agricultural insurance sampling terminal and receiving the target image data of the insured object from the agricultural insurance sampling terminal according to the selected sampling rule, includes the following steps: S51: Send the selected sampling rule to the agricultural insurance sampling terminal and collect the sampling identity information and sampling behavior data of the agricultural insurance sampling terminal; among which, the sampling behavior data includes sampling location information and image sampling parameters.

[0047] S52: Perform anomaly detection on the agricultural insurance sampling terminal based on the sampling identity information, sampling location information, and image sampling parameters to obtain anomaly detection information.

[0048] S53: Receive target image data fed back from the agricultural protection sampling terminal based on anomaly detection information.

[0049] In steps S51-S52, the compliance verification of the agricultural insurance sampling terminal mainly implements a triple verification mechanism of "identity credibility, location credibility, and image credibility" to ensure the credibility of the agricultural insurance sampling terminal before receiving the target image data uploaded by it. It should be noted that the triple sampling mechanism requires collecting the sampling identity information, sampling location information, and image sampling parameters of the agricultural insurance sampling terminal. The sampling identity information is used to verify "identity credibility," the sampling location information is used to verify "location credibility," and the image sampling parameters are used to verify "image credibility." Therefore, the sampling identity information, sampling location information, and image sampling parameters are used together to perform anomaly detection on the agricultural insurance sampling terminal. Anomaly detection determines anomalies in the three aspects of identity, location, and image, obtaining anomaly detection information characterizing anomalies in these three aspects.

[0050] In some embodiments, as disclosed above, verification of the agricultural insurance sampling terminal is required in terms of identity, location, and image. Please refer to... Figure 8 Step S52, which involves performing anomaly detection on the agricultural insurance sampling terminal based on the sampling identity information, sampling location information, and image sampling parameters to obtain anomaly detection information, includes the following steps: S531: Verify the identity of the agricultural insurance sampling terminal based on the preset identity information and the sampled identity information to obtain the identity verification information.

[0051] S532: Perform position verification on the agricultural protection sampling terminal based on preset position information and sampling position information to obtain position verification information.

[0052] S533: Perform sampling verification based on preset sampling parameters and image sampling parameters to obtain sampling verification information.

[0053] S534: Perform anomaly detection on the agricultural insurance sampling terminal based on identity verification information, location verification information, and sampling verification information to obtain anomaly detection information.

[0054] In step S531, the sampled identity information is the sampled object image. The sampled object features of the sampled object image are extracted, and the preset object features of the preset identity information are also extracted. The similarity between the sampled object features and the preset object features is calculated to obtain a feature similarity. If the feature similarity is greater than or equal to the preset similarity, the identity verification information indicates that the agricultural insurance sampling terminal's identity verification is successful; if the feature similarity is less than the preset similarity, the identity verification information indicates that the agricultural insurance sampling terminal's identity verification has failed. The sampled object features include at least one of the following: object action features, object facial texture features, object face features, object iris features, etc. The preset object features include specified action features, preset facial texture features, preset face features, and preset iris features. The identity verification information is determined by calculating the similarity between each sampled object feature and the preset object features one by one and then comparing the preset similarity.

[0055] For example, to verify the identity of sampling personnel at agricultural insurance sampling sites, they can be required to take photos or videos of themselves performing actions. An example analysis of the reflectivity of the sampling personnel's facial skin texture can be used to automatically block attacks such as photo manipulation, video synthesis, and 3D masks, achieving liveness detection of the sampling personnel. Furthermore, the current facial image of the sampling personnel is captured, and a reference facial image uploaded during the sampling personnel's real-name authentication is accessed via a cloud API interface. The similarity between the current facial image and the reference facial image is compared; if the similarity meets the standard, the sampling personnel's identity is deemed credible. Therefore, this embodiment incorporates liveness detection and facial authentication to improve the accuracy of identity verification information.

[0056] In step S532, the preset location information is the set location information of the above image sampling rules, which is the sampling location in the target protection area. By calculating the distance between the preset location information and the sampling location information, if the distance is within the preset tolerance range, the location verification information is determined to be location verification passed; if the distance is not within the preset tolerance range, the location verification information is determined to be location verification failed. Therefore, through location verification, suspicious sampling behaviors such as off-site check-in and fake check-in can be identified.

[0057] In step S533, the image sampling parameters are the parameters of the currently uploaded image from the agricultural protection sampling terminal, specifically including the current sampling time, current sampling latitude and longitude, current sampling device ID, and current operator ID. These image sampling parameters are embedded in the currently uploaded image as a watermark. Preset sampling parameters include preset sampling latitude and longitude, preset sampling device ID, and preset operator ID. Specifically, sampling verification information is determined by comparing the preset sampling parameters and the image sampling parameters. Therefore, by comparing the preset sampling parameters and the image sampling parameters, forged images can be identified, improving the accuracy of image credibility verification.

[0058] In step S534, in this embodiment, the identity verification information, location verification information, and sampling verification information are concatenated into anomaly detection information. It should be noted that if the identity verification information indicates successful identity verification, the location verification information indicates successful location verification, and the sampling verification information indicates successful sampling verification, then the anomaly detection information indicates that the agricultural insurance sampling terminal is normal, and the target image data sent by the agricultural insurance sampling terminal is received. If the identity verification information indicates failed identity verification, or the location verification information indicates failed location verification, or the sampling verification information indicates failed sampling verification, then the anomaly detection information indicates that the agricultural insurance sampling terminal has an anomaly.

[0059] The above solution, through the collaborative and comprehensive anomaly detection of identity, location, and parameters, achieves a significant improvement from single-dimensional verification to multi-dimensional cross-verification. It not only effectively solves the technical problem of the lack of compliance verification in traditional sampling inspections, but also significantly reduces the cost of manual review through an automated anomaly detection mechanism. It comprehensively enhances the authenticity, traceability, and management closed-loop capability of agricultural insurance verification data, providing a solid technical guarantee for accurate underwriting and accurate claims settlement.

[0060] In step S53, if the anomaly detection information indicates that the agricultural protection sampling terminal is normal, the target image data sent by the agricultural protection sampling terminal is received; if the anomaly detection information indicates that the agricultural protection sampling terminal is abnormal, the target image data uploaded by the agricultural protection sampling terminal is not directly received and further verification is required.

[0061] The above scheme, through the real-time transmission of identity, location, and parameter data, provides an objective and traceable original chain of evidence for subsequent verification. Furthermore, the conditional reception of target image data based on anomaly detection results forms an effective quality filtering mechanism. Only after the sampling process passes all verifications can the image be transmitted and stored, thus preventing image data based on false, invalid, or irregular sampling from flowing into the subsequent agricultural insurance underwriting engine at the data source.

[0062] In some embodiments, please refer to Figure 9 After step S533, the agricultural insurance underwriting method may further include the following steps: S71: If the anomaly detection information indicates that there is an anomaly at the agricultural insurance sampling terminal, the agricultural insurance sampling terminal is classified into anomalies based on the identity verification information, location verification information and sampling verification information to obtain the current anomaly category; S72: Select the target sampling review operation from the preset candidate sampling review operations based on the current anomaly category; S73: Perform target sampling and verification operations on the target insured area.

[0063] In steps S71-S72, as disclosed above, if the anomaly detection information indicates an anomaly at the agricultural insurance sampling terminal, the sampling terminal will be marked as an abnormal sampling terminal. Furthermore, the current anomaly category of the agricultural insurance sampling terminal needs to be determined by combining the identity verification information, location verification information, and sampling verification information. It should be noted that the current anomaly category includes identity anomaly category, location anomaly category, and sampling anomaly category. If it is an identity anomaly category, the target sampling verification operation is determined to be an identity verification operation; if it is a location anomaly category, the target sampling verification operation is determined to be a location verification operation; if it is a sampling anomaly category, the target sampling verification operation is determined to be a sampling verification operation.

[0064] In step S73, the target sampling verification operation is performed on the target insured area. This involves generating target sampling verification information based on the target sampling verification operation and sending the target sampling verification information to the management terminal of the agricultural insurance sampling terminal. The management terminal then completes the verification of the agricultural insurance sampling terminal in at least one aspect: identity, location, and sampling.

[0065] Furthermore, after the management system completes the review of the agricultural insurance sampling terminals, it sends the review results and determines whether to accept the target image data based on the results. If the review results indicate that the agricultural insurance sampling terminals are compliant, the target image data is accepted. If the review results indicate that the agricultural insurance sampling terminals are non-compliant, new agricultural insurance sampling terminals are selected, and the selected sampling rules are sent to the new agricultural insurance sampling terminals. This ensures the accuracy and efficiency of agricultural insurance underwriting and reduces the impact of sampling anomalies on the efficiency of agricultural insurance underwriting.

[0066] For example, if the identities of the sampling personnel at the agricultural insurance sampling terminal do not match, identity verification information is generated for the sampling personnel's manager, who then verifies the identity of the current sampling personnel to ensure the identity, location, and sampling accuracy of the sampling personnel, thereby improving the credibility of the target image data.

[0067] Through the above solution, this approach upgrades the traditional extensive anomaly handling model to a differentiated and refined governance model by establishing a complete technical chain of anomaly identification, precise classification, strategy matching, and targeted review. This significantly improves the targeting and efficiency of anomaly review and effectively ensures the authenticity, integrity, and compliance of image data during the agricultural insurance verification process.

[0068] S60: Underwrite the target insured area based on the underwriting model and target image data.

[0069] It should be noted that, based on the underwriting model, a target underwriting model is selected from a pre-set pool of candidate underwriting models. The agricultural insurance underwriting engine extracts reference image features of the insured objects from the target image data. Underwriting is then performed on the target insured area based on the target underwriting model and the reference image. Therefore, this embodiment achieves intelligent underwriting, eliminating the need for manual underwriting. Furthermore, the target image data accurately and truthfully represents the current status of the insured objects in the target insured area, improving underwriting accuracy and efficiency.

[0070] In addition, after the underwriting is completed, the agricultural insurance underwriting engine packages and automatically assembles the data generated in the above process to generate a standardized agricultural insurance acceptance report. The agricultural insurance acceptance report is then stored in the database to achieve full tracking of the agricultural insurance underwriting process and improve the traceability of agricultural insurance underwriting.

[0071] In summary, the sampling strategy in this application is upgraded from static designation to risk-adaptive designation, which significantly improves sample representativeness and data reliability. Compared with the traditional fixed ratio or manual designation mode, it constructs a risk profile of the target insured area by integrating multi-dimensional data such as historical compensation data, meteorological disaster data and remote sensing image data, and adaptively selects the sampling ratio and image sampling parameters based on the risk profile. This achieves an intelligent differential sampling mechanism of "strengthening sampling in high-risk areas and reasonable coverage in low-risk areas", which improves the representativeness of the sampled samples and makes reasonable use of sampling resources.

[0072] Furthermore, this embodiment establishes a triple intelligent verification mechanism encompassing identity verification, location verification, and sampling parameter verification. The cross-validation and collaborative analysis of the re-verified information automatically detects complex anomalies that cannot be identified by single-dimensional verification, curbing fraudulent operations at the source and effectively ensuring the authenticity and integrity of data throughout the entire verification chain. In addition, when anomalies are detected during the verification process, the system automatically categorizes them based on the three dimensions of identity, location, and image verification information, accurately distinguishing between different anomaly categories such as identity fraud, location deviation, and non-compliant parameters. It then adaptively matches differentiated review operations based on the anomaly category, achieving precise attribution and differentiated handling of anomaly events. Simultaneously, the entire process is traceable on the blockchain, supporting automatic anomaly alerts, intelligent task assignment, and full traceability of handling results, forming a complete management closed loop. This significantly shortens the anomaly response and handling cycle, effectively improving the efficiency and refined governance level of verification and acceptance.

[0073] As can be seen, in the above solution, for complex insured entities such as agricultural insurance businesses, the agricultural insurance underwriting engine first constructs the current risk level of the target insured area by combining multi-dimensional data such as historical claims data, meteorological disaster data, and remote sensing image data. Based on the current risk level, the regional location information of the target insured area, and the agricultural category of the insured, the sampling ratio and image sampling parameters are adaptively selected. Finally, according to the sampling ratio and image sampling parameters, the agricultural insurance sampling terminal is called to complete the target image data of the insured, and the underwriting of the target insured area is completed using the underwriting model and the target image data. Therefore, the data collection is targeted, and the underwriting is completed with more representative image data, which can improve the efficiency and accuracy of underwriting.

[0074] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0075] In one embodiment, an agricultural insurance underwriting device is provided, which corresponds one-to-one with the agricultural insurance underwriting methods described in the above embodiments. For example... Figure 10 As shown, the agricultural insurance underwriting device includes: a request receiving module 1001, a data acquisition module 1002, a risk assessment module 1003, a rule filtering module 1004, a sending and receiving module 1005, and an underwriting module 1006. Detailed descriptions of each functional module are as follows: The request receiving module 1001 is used to receive agricultural insurance underwriting requests for the target insured area; wherein, the target insured area is the area where the insured is located, and the agricultural insurance underwriting request includes: the insured's regional location information, agricultural category and insurance mode; Data acquisition module 1002 is used to collect meteorological disaster data, remote sensing image data and historical agricultural insurance compensation data of the target insured area based on regional location information and agricultural category; Risk assessment module 1003 is used to conduct risk assessment on the target insured area based on meteorological disaster data, remote sensing image data and historical agricultural insurance compensation data, and obtain the current risk level; The rule filtering module 1004 is used to filter out selected sampling rules from preset candidate sampling rules based on the current risk level, regional location information and agricultural category; wherein, the selected sampling rule represents the sampling ratio and image sampling requirements of the target insured area; The transceiver module 1005 is used to send the selected sampling rule to the agricultural insurance sampling terminal and receive the target image data of the insured object fed back by the agricultural insurance sampling terminal according to the selected sampling rule; The underwriting module 1006 is used to underwrite the target insured area based on the underwriting mode and target image data.

[0076] In one embodiment, the risk assessment module 1003 is specifically used for: Based on meteorological disaster data, risk prediction is performed on the target insured area to obtain the first risk value; Risk prediction is performed on the target insured area based on remote sensing image data to obtain a second risk value; Based on historical agricultural insurance claims data, risk prediction is performed on the target insured area to obtain a third risk value; The risk level of the target insured area is determined by assessing the risk based on the first, second, and third risk values.

[0077] In one embodiment, the risk assessment module 1003 is specifically used for: Assign a first weight to the first risk value, a second weight to the second risk value, and a third weight to the third risk value; The first risk value, the second risk value, and the third risk value are concatenated according to the first weight, the second weight, and the third weight to obtain the current risk value. The current risk level is determined based on the current risk value and the preset risk threshold.

[0078] In one embodiment, the rule filtering module 1004 is specifically used for: Select the chosen sampling ratio from the preset candidate sampling ratios based on the current risk level; Selected image sampling parameters are filtered from preset candidate image sampling parameters based on regional location information and agricultural category; The selected sampling ratio and the selected image sampling parameters are concatenated to obtain the selected sampling rule.

[0079] In one embodiment, the transceiver module 1005 is specifically used for: The selected sampling rules are sent to the agricultural insurance sampling terminal, and the sampling identity information and sampling behavior data of the agricultural insurance sampling terminal are collected; among which, the sampling behavior data includes sampling location information and image sampling parameters; Anomaly detection is performed on the agricultural protection sampling terminal based on the sampling identity information, sampling location information, and image sampling parameters to obtain anomaly detection information; Target image data is received from the agricultural insurance sampling terminal based on anomaly detection information.

[0080] In one embodiment, the transceiver module 1005 is specifically used for: The identity of the agricultural insurance sampling terminal is verified based on the preset identity information and the sampled identity information to obtain the identity verification information; The location of the agricultural protection sampling terminal is verified based on the preset location information and the sampling location information to obtain the location verification information. Sampling verification is performed based on preset sampling parameters and image sampling parameters to obtain sampling verification information; Anomaly detection is performed on the agricultural insurance sampling terminal based on identity verification information, location verification information, and sampling verification information to obtain anomaly detection information.

[0081] In one embodiment, the agricultural insurance underwriting device further includes a verification module, specifically used for: If the anomaly detection information indicates that there is an anomaly at the agricultural insurance sampling terminal, the agricultural insurance sampling terminal is classified into anomalies based on identity verification information, location verification information, and sampling verification information to obtain the current anomaly category. The target sampling and review operation is selected from the preset candidate sampling and review operations based on the current anomaly category; Perform target sampling and verification operations on the target insured area.

[0082] This invention provides an agricultural insurance underwriting device. First, it receives an underwriting request containing regional location information, agricultural category, and underwriting mode. Then, it collects three types of multi-source heterogeneous data: meteorological disaster data, remote sensing imagery, and historical claims data. Based on the dynamic risk level obtained from the fusion assessment, it adaptively selects sampling rules that match regional characteristics and agricultural type. These rules are then precisely sent to the sampling end, which receives the target image data in return. Finally, the underwriting determination is completed by combining the underwriting mode. Therefore, this embodiment achieves the scientific quantification of land risk through multi-source data fusion assessment and realizes differentiated and precise sampling through risk-driven dynamic sampling rules, improving underwriting efficiency and accuracy.

[0083] Specific limitations regarding the agricultural insurance underwriting device can be found in the limitations of the intelligent question-and-answer method described above, and will not be repeated here. Each module in the aforementioned agricultural insurance underwriting device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the corresponding operations of each module.

[0084] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 11As shown, the computer device includes a processor, memory, network interface, and database connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile and / or volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The network interface is used to communicate with external clients via a network connection. When the computer program is executed by the processor, it implements the functions or steps of a server-side agricultural insurance underwriting method.

[0085] In one embodiment, a computer device is provided, which may be a client, and its internal structure diagram may be as follows: Figure 12 As shown, the computer device includes a processor, memory, network interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The network interface is used to communicate with an external server via a network connection. When the computer program is executed by the processor, it implements the functions or steps of a client-side agricultural insurance underwriting method.

[0086] In one embodiment, a computer device is provided, 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 perform the following steps: Receive agricultural insurance underwriting requests from the target insured area; where the target insured area is the region where the insured is located, and the agricultural insurance underwriting request includes: the insured's regional location information, agricultural category, and insurance model; Based on regional location information and agricultural category, meteorological disaster data, remote sensing image data, and historical agricultural insurance compensation data of the target insured areas are collected. A risk assessment is conducted on the target insured area based on meteorological disaster data, remote sensing image data, and historical agricultural insurance compensation data to determine the current risk level; Based on the current risk level, regional location information, and agricultural category, a selected sampling rule is selected from the preset candidate sampling rules; whereby the selected sampling rule represents the sampling ratio and image sampling requirements of the target insured area; Send the selected sampling rule to the agricultural insurance sampling terminal, and receive the target image data of the insured object from the agricultural insurance sampling terminal according to the selected sampling rule; Underwriting is conducted on the target insured area based on the underwriting model and target image data.

[0087] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor: Receive agricultural insurance underwriting requests from the target insured area; where the target insured area is the region where the insured is located, and the agricultural insurance underwriting request includes: the insured's regional location information, agricultural category, and insurance model; Based on regional location information and agricultural category, meteorological disaster data, remote sensing image data, and historical agricultural insurance compensation data of the target insured areas are collected. A risk assessment is conducted on the target insured area based on meteorological disaster data, remote sensing image data, and historical agricultural insurance compensation data to determine the current risk level; Based on the current risk level, regional location information, and agricultural category, a selected sampling rule is selected from the preset candidate sampling rules; whereby the selected sampling rule represents the sampling ratio and image sampling requirements of the target insured area; Send the selected sampling rule to the agricultural insurance sampling terminal, and receive the target image data of the insured object from the agricultural insurance sampling terminal according to the selected sampling rule; Underwriting is conducted on the target insured area based on the underwriting model and target image data.

[0088] It should be noted that the functions or steps that can be implemented by the computer-readable storage medium or computer device described above can be referred to the relevant descriptions on the server side and client side in the foregoing method embodiments. To avoid repetition, they will not be described one by one here.

[0089] Those skilled in the art will understand that all or part of the processes in the methods of 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 of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0090] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0091] It should be noted that any AI models, software tools, or components not belonging to this company appearing in the embodiments of this application are merely illustrative examples and do not represent actual use. All user personal information involved in the embodiments of this application has been authorized (with the knowledge and consent) by the relevant parties or has been fully authorized by all parties, and the executing entity may obtain it through various legal and compliant means. The collection, storage, use, processing, transmission, provision, and disclosure of the information, data, and signals involved all comply with relevant laws and regulations and do not violate public order and good morals.

[0092] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for agricultural insurance underwriting, characterized in that, include: Receive agricultural insurance underwriting requests for the target insured area; wherein, the target insured area is the area where the insured is located, and the agricultural insurance underwriting request includes: the insured's regional location information, agricultural category, and insurance mode; Based on the regional location information and the agricultural category, meteorological disaster data, remote sensing image data, and historical agricultural insurance compensation data of the target insured area are collected. A risk assessment is conducted on the target insured area based on the meteorological disaster data, the remote sensing image data, and the historical agricultural insurance compensation data to obtain the current risk level; Based on the current risk level, the regional location information, and the agricultural category, a selected sampling rule is selected from the preset candidate sampling rules; wherein, the selected sampling rule represents the sampling ratio and image sampling requirements of the target insured area; The selected sampling rule is sent to the agricultural insurance sampling terminal, and the target image data of the insured object is received from the agricultural insurance sampling terminal according to the selected sampling rule. The target insured area is underwritten based on the underwriting model and the target image data.

2. The agricultural insurance underwriting method as described in claim 1, characterized in that, The risk assessment of the target insured area based on the meteorological disaster data, the remote sensing image data, and the historical agricultural insurance compensation data, to obtain the current risk level, includes: Based on the meteorological disaster data, a risk prediction is made for the target insured area to obtain a first risk value; Based on the remote sensing image data, a risk prediction is performed on the target insured area to obtain a second risk value; Based on the historical agricultural insurance claims data, a risk prediction is made for the target insured area to obtain a third risk value; The target insured area is assessed for risk based on the first risk value, the second risk value, and the third risk value to obtain the current risk level.

3. The agricultural insurance underwriting method as described in claim 2, characterized in that, The step of assessing the risk of the target insured area based on the first risk value, the second risk value, and the third risk value to obtain the current risk level includes: Assign a first weight to the first risk value, a second weight to the second risk value, and a third weight to the third risk value; The first risk value, the second risk value, and the third risk value are concatenated according to the first weight, the second weight, and the third weight to obtain the current risk value; The current risk level is determined based on the current risk value and the preset risk threshold.

4. The agricultural insurance underwriting method as described in claim 1, characterized in that, The step of selecting a sampling rule from preset candidate sampling rules based on the current risk level, the regional location information, and the agricultural category includes: The selected sampling ratio is selected from the preset candidate sampling ratios based on the current risk level. Selected image sampling parameters are selected from preset candidate image sampling parameters based on the regional location information and the agricultural category; The selected sampling ratio and the selected image sampling parameters are concatenated to obtain the selected sampling rule.

5. The agricultural insurance underwriting method as described in any one of claims 1 to 4, characterized in that, The step of sending the selected sampling rule to the agricultural insurance sampling terminal and receiving the target image data of the insured object from the agricultural insurance sampling terminal according to the selected sampling rule includes: The selected sampling rule is sent to the agricultural insurance sampling terminal, and the sampling identity information and sampling behavior data of the agricultural insurance sampling terminal are collected; wherein, the sampling behavior data includes sampling location information and image sampling parameters; Anomaly detection is performed on the agricultural insurance sampling terminal based on the sampling identity information, the sampling location information, and the image sampling parameters to obtain anomaly detection information; The target image data fed back by the agricultural protection sampling terminal is received based on the anomaly detection information.

6. The agricultural insurance underwriting method as described in claim 5, characterized in that, The step of performing anomaly detection on the agricultural insurance sampling terminal based on the sampling identity information, the sampling location information, and the image sampling parameters to obtain anomaly detection information includes: The identity of the agricultural insurance sampling terminal is verified based on the preset identity information and the sampling identity information to obtain identity verification information. The location of the agricultural protection sampling terminal is verified based on the preset location information and the sampling location information to obtain location verification information; Sampling verification is performed based on preset sampling parameters and the image sampling parameters to obtain sampling verification information; Anomaly detection is performed on the agricultural insurance sampling terminal based on the identity verification information, the location verification information, and the sampling verification information to obtain the anomaly detection information.

7. The agricultural insurance underwriting method as described in claim 6, characterized in that, After performing anomaly detection on the agricultural insurance sampling terminal based on the identity verification information, the location verification information, and the sampling verification information to obtain the anomaly detection information, the method further includes: If the anomaly detection information indicates that there is an anomaly at the agricultural insurance sampling terminal, the agricultural insurance sampling terminal is classified into anomalies based on the identity verification information, the location verification information, and the sampling verification information to obtain the current anomaly category. The target sampling review operation is selected from the preset candidate sampling review operations based on the current anomaly category; Perform the target sampling and verification operation on the target insured area.

8. An agricultural insurance underwriting device, characterized in that, include: The request receiving module is used to receive agricultural insurance underwriting requests for the target insured area; wherein, the target insured area is the area where the insured is located, and the agricultural insurance underwriting request includes: the insured's regional location information, agricultural category, and insurance mode; The data acquisition module is used to collect meteorological disaster data, remote sensing image data and historical agricultural insurance compensation data of the target insured area based on the regional location information and the agricultural category; The risk assessment module is used to conduct a risk assessment of the target insured area based on the meteorological disaster data, the remote sensing image data, and the historical agricultural insurance compensation data, and to obtain the current risk level. The rule filtering module is used to filter out selected sampling rules from preset candidate sampling rules based on the current risk level, the regional location information, and the agricultural category; wherein, the selected sampling rule represents the sampling ratio and image sampling requirements of the target insured area; The transceiver module is used to send the selected sampling rule to the agricultural insurance sampling terminal and receive the target image data of the insured object fed back by the agricultural insurance sampling terminal according to the selected sampling rule; The underwriting module is used to underwrite the target insured area based on the underwriting mode and the target image data.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the agricultural insurance underwriting method as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the agricultural insurance underwriting method as described in any one of claims 1 to 7.

Images