Loan request processing method and device based on tree and electronic equipment
By combining the target weather conditions and pre-trained prediction models in the tree image evaluation, the growth and health status of trees are determined, and the inaccurate assessment of tree value caused by weather factors is solved, and the accuracy of loan request processing is improved.
Patent Information
- Application Number
- CN202410682744.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-05-29
- Publication Date
- 2025-01-17
AI Technical Summary
In the prior art, weather conditions have inaccurate assessment of tree images, resulting in inaccurate processing of loan requests, and lack of effective solutions.
By obtaining the remote sensing image of the target tree under the target weather, extracting image features, and determining the target prediction model based on the pre-trained prediction models of multiple weather types, converting the image features into vectors, and combining prior knowledge to determine the growth potential and health status of the trees. information, and then process loan requests.
It improves the accuracy of tree growth and health assessment, enhances the accuracy of loan request processing, and solves the problem of inaccurate image value assessment caused by weather factors.
Smart Images
Figure CN119323462A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of artificial intelligence, and more specifically, to a tree-based loan request processing method, device and electronic device. Background Art
[0002] In the prior art, tree growers can use trees as collateral to apply for loans from banks and other financial institutions. Banks and other financial institutions can assess the value of the trees based on images taken of the trees, and then determine the maximum loan amount for the trees and whether to approve the loan request.
[0003] However, due to the differences in images of trees taken under different weather conditions, it is easy to cause inaccurate assessment of tree value based on the photographed tree images, which in turn affects the accuracy of processing loan requests.
[0004] To address the above-mentioned problems, no effective solution has been proposed yet. Summary of the invention
[0005] The present application provides a tree-based loan request processing method, device and electronic device to at least solve the technical problem in the prior art that weather factors lead to inaccurate tree value assessment based on photographed tree images, thereby causing inaccurate loan request processing.
[0006] According to one aspect of the present application, a tree-based loan request processing method is provided, comprising: obtaining a target remote sensing image taken of a target tree under target weather conditions; extracting target image features of the target tree in the target remote sensing image; determining at least one target prediction model from X pre-trained prediction models according to the weather type of the target weather, wherein the X prediction models correspond to X weather types, and X is an integer greater than 1; converting the target image features into a target vector through the target prediction model, and determining the growth potential information and health status information of the target tree according to the target vector based on the prior knowledge learned during the model training process; and processing the loan request with the target tree as collateral according to the growth potential information and health status information of the target tree.
[0007] Optionally, the tree-based loan request processing method further includes: when the weather type of the target weather is the i-th weather type among X weather types, using the prediction model corresponding to the i-th weather type as the target prediction model, wherein i is a positive integer less than or equal to X.
[0008] Optionally, the tree-based loan request processing method also includes: when the weather type of the target weather is a mixed weather type consisting of M weather types out of X weather types, using M prediction models corresponding to the M weather types as M target prediction models, wherein M is an integer less than or equal to X and greater than 1.
[0009] Optionally, the tree-based loan request processing method also includes: after taking the M prediction models corresponding to the M weather types as M target prediction models, converting the target image features into target vectors through each of the M target prediction models; determining the health score and growth score of the target tree according to the target vector through the prior knowledge learned by each target prediction model during the model training process, and obtaining M health scores and M growth scores, wherein the M health scores are respectively determined by the M prediction models, and the M growth scores are respectively determined by the M prediction models; obtaining a target health score by calculating the statistical data of the M health scores, and characterizing the health information of the target tree by the target health score; obtaining a target growth score by calculating the statistical data of the M growth scores, and characterizing the growth information of the target tree by the target growth score.
[0010] Optionally, the tree-based loan request processing method also includes: performing preliminary feature selection on the target remote sensing image to obtain N image features, where N is an integer greater than 1; calculating the regression coefficient of each of the N image features, where the regression coefficient of each image feature is used to characterize the degree of influence of the image feature on determining the health status information and growth information of the target tree; and determining the target image feature from the N image features based on the regression coefficient.
[0011] Optionally, the tree-based loan request processing method also includes: determining the value of the target tree based on the growth information and health status information of the target tree; approving the loan request when the value of the target tree is greater than or equal to the loan funds applied for in the loan request; generating risk warning information when the value of the target tree is less than the loan funds applied for in the loan request.
[0012] Optionally, the tree-based loan request processing method further includes: obtaining growth information and health status information of reference trees of the same type as the target tree at Y historical moments, where Y is an integer greater than 1; obtaining Z historical images taken for the reference tree at Y historical moments, wherein each of the Y historical moments corresponds to at least one historical image taken, and Z is an integer greater than or equal to Y; dividing the Z historical images into X image sets according to the weather type of the area where the target tree is located at each historical moment, wherein the historical images in each image set correspond to the same weather type; and obtaining X prediction models by training based on the X image sets, wherein the training data of each prediction model includes historical images in an image set and growth information and health status information of the reference tree when the historical images in the image set are taken.
[0013] Optionally, the statistical data is any one of a maximum value, a minimum value, a mean value, a median, and a mode.
[0014] According to another aspect of the present application, a tree-based loan request processing device is also provided, wherein the tree-based loan request processing device includes: an acquisition unit, used to acquire a target remote sensing image taken of a target tree under target weather; a feature extraction unit, used to extract target image features of the target tree in the target remote sensing image; a first determination unit, used to determine at least one target prediction model from X pre-trained prediction models according to the weather type of the target weather, wherein the X prediction models correspond to X weather types, and X is an integer greater than 1; a second determination unit, used to convert the target image features into a target vector through the target prediction model, and determine the growth potential information and health status information of the target tree according to the target vector based on the prior knowledge learned during the model training process; and a processing unit, used to process the loan request with the target tree as collateral according to the growth potential information and health status information of the target tree.
[0015] According to another aspect of the present application, a computer-readable storage medium is provided, wherein a computer program is stored in the computer-readable storage medium, wherein the computer program is configured to execute any of the tree-based loan request processing methods described above when run.
[0016] According to another aspect of the present application, an electronic device is also provided, wherein the electronic device includes one or more processors; a storage device for storing one or more programs, which, when the one or more programs are executed by the one or more processors, enables the one or more processors to implement the operation of the programs, wherein the programs are configured to execute any of the above-mentioned tree-based loan request processing methods when running.
[0017] In the present application, at least one target prediction model is determined from X pre-trained prediction models according to the weather type of the target weather. First, a target remote sensing image taken for a target tree under the target weather is obtained, and then the target image features of the target tree in the target remote sensing image are extracted. Subsequently, at least one target prediction model is determined from X pre-trained prediction models according to the weather type of the target weather, wherein the X prediction models correspond to X weather types, and X is an integer greater than 1. The target image features are converted into a target vector through the target prediction model, and the growth potential information and health status information of the target tree are determined according to the target vector based on the prior knowledge learned during the model training process. Finally, the loan request with the target tree as collateral is processed according to the growth potential information and health status information of the target tree.
[0018] From the above content, it can be seen that in the technical solution of the present application, by combining the weather conditions when the target remote sensing image was taken, the target prediction model is selectively used to determine the growth information and health status information of the target trees, thereby improving the accuracy of the assessment of the growth and health conditions of the target trees, and further improving the accuracy of processing loan requests with the target trees as collateral.
[0019] It can be seen that the technical solution of the present application achieves the purpose of determining the health status and growth of trees based on tree images based on the weather factors when the tree images were taken, thereby achieving the technical effect of improving the accuracy of the assessment of the growth and health conditions of the target trees, and further solving the technical problem in the prior art of inaccurate tree value assessment based on the taken tree images due to weather factors, which in turn causes inaccurate loan request processing. BRIEF DESCRIPTION OF THE DRAWINGS
[0020] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:
[0021] Figure 1 is a flow chart of an optional tree-based loan request processing method according to an embodiment of the present application;
[0022] Figure 2 This is an optional flow chart of determining the growth information and health status information of target trees based on mixed weather types according to an embodiment of the present application;
[0023] Figure 3 is a training flow chart of an optional prediction model according to an embodiment of the present application;
[0024] Figure 4is a schematic diagram of an optional tree-based loan request processing device according to an embodiment of the present application;
[0025] Figure 5 It is a schematic diagram of an electronic device provided according to an embodiment of the present application. DETAILED DESCRIPTION
[0026] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.
[0027] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.
[0028] It should also be noted that the information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) collected by this application are information and data authorized by the user or fully authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of relevant data are in compliance with the relevant laws, regulations and standards of the relevant regions, necessary confidentiality measures are taken, and public order and good customs are not violated, and corresponding operation portals are provided for users to choose to authorize or refuse. For example, an interface is set up between this system and relevant users or organizations. Before obtaining relevant information, it is necessary to send an acquisition request to the aforementioned user or organization through the interface, and obtain relevant information after receiving the consent information fed back by the aforementioned user or organization.
[0029] In addition, in this application, when collecting and analyzing customer information, the corresponding operation entrance is provided for users to choose to agree or reject the automated decision-making results; if the user chooses to reject, the expert decision-making process will be entered.
[0030] According to an embodiment of the present application, an embodiment of a tree-based loan request processing method is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.
[0031] Figure 1 is a flowchart of an optional tree-based loan request processing method according to an embodiment of the present application, such as Figure 1 As shown, the method comprises the following steps:
[0032] Step S101, obtaining a target remote sensing image taken of a target tree under target weather conditions.
[0033] In an optional embodiment, a loan request processing system can serve as the executor of the tree-based loan request processing method in the embodiment of the present application, wherein the loan request processing system can be a software system or an embedded system combining software and hardware.
[0034] Optionally, the target trees may be various types of trees, such as peach trees, apricot trees, pear trees, etc. The loan request processing system may control the satellite or drone to capture target remote sensing images of the target trees by sending image acquisition instructions to the satellite or drone.
[0035] It should be noted that there can be multiple target remote sensing images. In addition, when taking different target remote sensing images, the weather conditions in the area where the target trees are located may be different, which will affect the differences in the target remote sensing images taken by drones or satellites of the target trees. Such differences include but are not limited to differences in image quality, image content, image color, image contrast, image brightness, etc.
[0036] Step S102, extracting target image features of target trees in the target remote sensing image.
[0037] Optionally, the target image features include, but are not limited to, the contour features, edge features, color features, shape features, and brightness features of the target tree in the target remote sensing image. Furthermore, the target image features may also include the image features of the fruit, the image features of the leaves, and the image features of the branches of the target tree in the target remote sensing image, wherein the image features of the fruit include, but are not limited to, the number of fruits, the size of fruits, the color of fruits, the integrity of fruits, etc., the image features of leaves include, but are not limited to, the size of leaves, the density of leaves, the integrity of leaves, the color of leaves, and the image features of branches include, but are not limited to, the number of branches, the length of branches, and the thickness of branches.
[0038] Step S103: determining at least one target prediction model from the pre-trained X prediction models according to the weather type of the target weather.
[0039] In step S103, X prediction models correspond to X weather types, where X is an integer greater than 1.
[0040] Optionally, the X weather types include but are not limited to sunny, overcast, cloudy, foggy, rainy, snowy, haze, dusty, etc. Among them, one prediction model in the X prediction models corresponds to one weather type in the X weather types, and different prediction models correspond to different weather types.
[0041] Step S104, converting the target image features into a target vector through the target prediction model, and determining the growth potential information and health status information of the target tree according to the target vector based on the prior knowledge learned during the model training process.
[0042] Optionally, the target prediction model can be a deep learning neural network model, which can first convert the target image features into the form of a target vector, and then determine the growth information and health status information of the target tree based on the target vector using the prior knowledge learned during the model training process.
[0043] Step S105, processing the loan request with the target tree as collateral according to the growth information and health status information of the target tree.
[0044] Optionally, the loan request processing system may determine the value of the target tree based on the growth information and health status information of the target tree, and then determine the processing operation to be performed on the loan request based on the value of the target tree and the loan funds applied for in the loan request.
[0045] Based on the contents of the above steps S101 to S105, it can be known that in the present application, at least one target prediction model is determined from X pre-trained prediction models according to the weather type of the target weather. First, a target remote sensing image taken for the target tree under the target weather is obtained, and then the target image features of the target tree in the target remote sensing image are extracted. Subsequently, at least one target prediction model is determined from X pre-trained prediction models according to the weather type of the target weather, wherein the X prediction models correspond to X weather types, and X is an integer greater than 1. The target image features are converted into a target vector through the target prediction model, and the growth potential information and health status information of the target tree are determined according to the target vector based on the prior knowledge learned during the model training process. Finally, the loan request with the target tree as collateral is processed according to the growth potential information and health status information of the target tree.
[0046] From the above content, it can be seen that in the technical solution of the present application, by combining the weather conditions when the target remote sensing image was taken, the target prediction model is selectively used to determine the growth information and health status information of the target trees, thereby improving the accuracy of the assessment of the growth and health conditions of the target trees, and further improving the accuracy of processing loan requests with the target trees as collateral.
[0047] It can be seen that the technical solution of the present application achieves the purpose of determining the health status and growth of trees based on tree images based on the weather factors when the tree images were taken, thereby achieving the technical effect of improving the accuracy of the assessment of the growth and health conditions of the target trees, and further solving the technical problem in the prior art of inaccurate tree value assessment based on the taken tree images due to weather factors, which in turn causes inaccurate loan request processing.
[0048] In an optional embodiment, when the weather type of the target weather is the i-th weather type among X weather types, the prediction model corresponding to the i-th weather type is used as the target prediction model, where i is a positive integer less than or equal to X.
[0049] Optionally, if the target weather is sunny, the prediction model corresponding to sunny days is selected as the target prediction model; if the target weather is rainy days, the prediction model corresponding to rainy days is selected as the target prediction model.
[0050] It should be noted that by selecting the corresponding prediction model based on the weather type of the target weather as the target prediction model, it can be ensured that the prediction model that best matches the target remote sensing image is selected for subsequent processing of the target remote sensing image, wherein the best match means referring to the weather conditions when the target remote sensing image was taken, and selecting the prediction model corresponding to the same weather type as the prediction model that best matches the target remote sensing image, thereby avoiding the problem of inaccurate tree value assessment based on the captured tree images caused by weather factors.
[0051] In an optional embodiment, when the weather type of the target weather is a mixed weather type consisting of M weather types out of X weather types, M prediction models corresponding to the M weather types are used as M target prediction models, where M is an integer less than or equal to X and greater than 1.
[0052] Optionally, if the target weather is a mixture of sunny and dusty weather, it is necessary to simultaneously determine the prediction model corresponding to sunny weather and the prediction model corresponding to dusty weather as the target prediction model. It should be noted that weather is a variable factor at any time. For example, at time T1, it may be sunny but accompanied by dusty weather; at time T2, it may be rainy but accompanied by foggy weather.
[0053] For mixed weather types, the present application can decompose the mixed weather types into various single weather types, and then determine the prediction model corresponding to each single weather type as the target prediction model, thereby improving the accuracy of selecting the target prediction model.
[0054] In an optional embodiment, after taking the M prediction models corresponding to the M weather types as the M target prediction models, the loan request processing system can convert the target image features into target vectors through each of the M target prediction models, and then determine the health score and growth score of the target tree according to the target vectors through the prior knowledge learned by each target prediction model during the model training process, and obtain M health scores and M growth scores, wherein the M health scores are determined by the M prediction models respectively, and the M growth scores are determined by the M prediction models respectively. Subsequently, the loan request processing system obtains the target health score by calculating the statistical data of the M health scores, and characterizes the health information of the target tree through the target health score; obtains the target growth score by calculating the statistical data of the M growth scores, and characterizes the growth information of the target tree through the target growth score.
[0055] Optionally, the statistical data may be any one of a maximum value, a minimum value, a mean value, a median value, and a mode value.
[0056] Figure 2 is an optional flow chart of determining the growth information and health status information of target trees based on mixed weather types according to an embodiment of the present application, such as Figure 2 As shown, when the target weather is a mixed weather type where sunny days and dusty days coexist, the loan request processing system simultaneously determines the prediction model corresponding to sunny days and the prediction model corresponding to dusty days as the target prediction model. Assume that the loan request processing system uses the prediction model corresponding to sunny days to determine that the health score of the target tree is T1, and the growth score is F1; and uses the prediction model corresponding to dusty days to determine that the health score of the target tree is T2, and the growth score is F2. Finally, the loan request processing system can combine T1 and T2 to determine the target health score, and combine F1 and F2 to determine the target growth score. For example, when the statistical data is the maximum value, the maximum value of T1 and T2 is selected as the target health score, and the maximum value of F1 and F2 is selected as the target growth score.
[0057] It should be noted that for mixed weather types, different weather types may give different health scores and / or growth scores. The present application performs mathematical statistics on the health scores and growth scores determined by different target prediction models, so as to comprehensively and comprehensively balance the health scores and growth scores determined by different target prediction models, thereby ensuring that accurate health status information and growth information of the target trees are obtained.
[0058] It should also be noted that when calculating statistical data, a weight can also be set for each target prediction model based on the weather type corresponding to each target prediction model. For example, if the weather type corresponding to a target prediction model has a greater degree of interference with image shooting, a smaller weight is set for the target prediction model. For example, since the image quality taken on a sunny day is better and can more accurately predict the health status and growth status of the target trees, while the image quality taken on a dusty day is poor, the prediction accuracy of the health status and growth status of the target trees is poor, so a larger weight can be set for the prediction model corresponding to a sunny day, and a smaller weight can be set for the prediction model corresponding to a dusty day.
[0059] In an optional embodiment, the loan request processing system may also perform preliminary feature selection on the target remote sensing image to obtain N image features, where N is an integer greater than 1. Then, the loan request processing system calculates the regression coefficient of each image feature in the N image features, where the regression coefficient of each image feature is used to characterize the degree of influence of the image feature on determining the health status information and growth information of the target tree. Finally, the loan request processing system determines the target image feature from the N image features based on the regression coefficient.
[0060] Optionally, the present application provides a feature processing algorithm to calculate the regression coefficient of each image feature, wherein the regression coefficient of the image feature is used to characterize the degree of influence of the image feature on determining the health status information and growth potential information of the target tree.
[0061] Optionally, the basic principle of the above feature processing algorithm is to minimize the residual sum of squares under the constraint that the regression coefficient is less than or equal to a certain threshold, and the regression coefficient of the image features with less influence is compressed to zero, so that the features with less influence on determining the health status information and growth information of the target tree can be excluded. The mathematical expression of minimizing the residual sum of squares problem is as follows:
[0062]
[0063]
[0064] Among them, x ij To determine the characteristics of the health status and growth information of the target trees; i Information about the health status and growth of trees; j is the regression coefficient of the jth image feature; the threshold t is the value of β jThe normal form penalty has a value range of approximately [0, +∞]. When the value is small, the regression coefficients of some features that have little influence on determining the health status information and growth information of the target trees will be compressed to 0. By discarding the features whose regression coefficients are compressed to 0, more important target image features can be selected from multiple features.
[0065] It should be noted that the regression coefficient is introduced to evaluate the importance of each image feature, so that when there are a large number of image features, more important image features can be screened as target image features, and unimportant features can be ignored as redundant features, thereby improving the efficiency of subsequent feature processing and avoiding the interference of redundant features.
[0066] In an optional embodiment, the loan request processing system can determine the value of the target tree based on the growth information and health status information of the target tree. If the value of the target tree is greater than or equal to the loan amount applied for in the loan request, the loan request is approved; if the value of the target tree is less than the loan amount applied for in the loan request, risk warning information is generated.
[0067] Optionally, the present application first determines the value of the target trees based on the growth information and health status information of the target trees, and then determines how to process the loan request based on the relationship between the value and the amount of the loan application funds, thereby minimizing the risk of the loan funds not being recovered.
[0068] In an optional embodiment, Figure 3 is a training flow chart of an optional prediction model according to an embodiment of the present application, such as Figure 3 As shown, the following steps are included:
[0069] Step S301, obtaining growth information and health status information of reference trees of the same type as the target tree at Y historical moments, where Y is an integer greater than 1.
[0070] Optionally, using the growth information and health status information of reference trees of the same type as the target tree at Y historical moments as training data can achieve the purpose of maximizing the control variables and avoid the problem of poor model prediction accuracy due to different tree types.
[0071] Step S302: Obtain Z historical images taken for the reference tree at Y historical moments, wherein each of the Y historical moments corresponds to at least one historical image, and Z is an integer greater than or equal to Y.
[0072] Optionally, at least one historical image may be continuously captured at each historical moment, so Z is an integer greater than or equal to Y.
[0073] Step S303 , dividing the Z historical images into X image sets according to the weather type of the area where the target tree is located at each historical moment, wherein the historical images in each image set correspond to the same weather type.
[0074] Optionally, assuming that there are 8 weather types in total, the Z historical images are divided into 8 image sets, and all the historical images included in each image set correspond to the same weather type.
[0075] Step S304, obtaining X prediction models through training based on the X image sets, wherein the training data of each prediction model includes a historical image in an image set and growth potential information and health status information of reference trees when the historical image in the image set is captured.
[0076] Optionally, assuming that image set 1 includes historical image page-1, historical image page-2 and historical image page-3, wherein these three historical images are all taken on sunny days, the loan request processing system may use historical image page-1, historical image page-2 and historical image page-3 as training images, and refer to the growth information and health status information of the trees when taking historical image page-1, the growth information and health status information of the trees when taking historical image page-2, and the growth information and health status information of the trees when taking historical image page-3 as training labels. After inputting the training images and training labels into the deep learning neural network and performing multiple iterative training, a prediction model corresponding to a sunny day can be trained.
[0077] It should be noted that by training a prediction model based on a weather type, after obtaining the target remote sensing image of the target tree taken under the target weather, it is possible to quickly match it to the prediction model corresponding to the target weather, thereby solving the problem in the prior art of inaccurate tree value assessment based on the captured tree images due to weather factors.
[0078] From the above content, it can be seen that in the technical solution of the present application, by combining the weather conditions when the target remote sensing image was taken, the target prediction model is selectively used to determine the growth information and health status information of the target trees, thereby improving the accuracy of the assessment of the growth and health conditions of the target trees, and further improving the accuracy of processing loan requests with the target trees as collateral.
[0079] It can be seen that the technical solution of the present application achieves the purpose of determining the health status and growth of trees based on tree images based on the weather factors when the tree images were taken, thereby achieving the technical effect of improving the accuracy of the assessment of the growth and health conditions of the target trees, and further solving the technical problem in the prior art of inaccurate tree value assessment based on the taken tree images due to weather factors, which in turn causes inaccurate loan request processing.
[0080] According to an embodiment of the present application, an embodiment of a tree-based loan request processing device is also provided. Figure 4 is a schematic diagram of an optional tree-based loan request processing device according to an embodiment of the present application, such as Figure 4 As shown, the tree-based loan request processing device includes: an acquisition unit 401 , a feature extraction unit 402 , a first determination unit 403 , a second determination unit 404 , and a processing unit 405 .
[0081] Optionally, an acquisition unit 401 is used to acquire a target remote sensing image taken of a target tree under target weather conditions; a feature extraction unit 402 is used to extract target image features of the target tree in the target remote sensing image; a first determination unit 403 is used to determine at least one target prediction model from X pre-trained prediction models according to the weather type of the target weather, wherein the X prediction models correspond to X weather types, and X is an integer greater than 1; a second determination unit 404 is used to convert the target image features into a target vector through the target prediction model, and determine the growth potential information and health status information of the target tree according to the target vector based on the prior knowledge learned during the model training process; and a processing unit 405 is used to process a loan request with the target tree as collateral according to the growth potential information and health status information of the target tree.
[0082] From the above content, it can be seen that in the technical solution of the present application, by combining the weather conditions when the target remote sensing image was taken, the target prediction model is selectively used to determine the growth information and health status information of the target trees, thereby improving the accuracy of the assessment of the growth and health conditions of the target trees, and further improving the accuracy of processing loan requests with the target trees as collateral.
[0083] It can be seen that the technical solution of the present application achieves the purpose of determining the health status and growth of trees based on tree images based on the weather factors when the tree images were taken, thereby achieving the technical effect of improving the accuracy of the assessment of the growth and health conditions of the target trees, and further solving the technical problem in the prior art of inaccurate tree value assessment based on the taken tree images due to weather factors, which in turn causes inaccurate loan request processing.
[0084] Optionally, the first determination unit 403 includes: a first processing subunit, used to use the prediction model corresponding to the i-th weather type as the target prediction model when the weather type of the target weather is the i-th weather type among X weather types, where i is a positive integer less than or equal to X.
[0085] Optionally, the first determination unit 403 includes: a second processing subunit, used to use M prediction models corresponding to the M weather types as M target prediction models when the weather type of the target weather is a mixed weather type composed of M weather types out of X weather types, wherein M is an integer less than or equal to X and greater than 1.
[0086] Optionally, the second determination unit 404 includes: a third processing subunit, a fourth processing subunit, a fifth processing subunit, and a sixth processing subunit. The third processing subunit is used to convert the target image features into target vectors through each of the M target prediction models after taking the M prediction models corresponding to the M weather types as M target prediction models; the fourth processing subunit is used to determine the health score and growth score of the target tree according to the target vector based on the prior knowledge learned by each target prediction model in the model training process, and obtain M health scores and M growth scores, wherein the M health scores are determined by the M prediction models respectively, and the M growth scores are determined by the M prediction models respectively; the fifth processing subunit is used to obtain the target health score by calculating the statistical data of the M health scores, and to characterize the health information of the target tree by the target health score; the sixth processing subunit is used to obtain the target growth score by calculating the statistical data of the M growth scores, and to characterize the growth information of the target tree by the target growth score.
[0087] Optionally, the feature extraction unit 402 includes: a seventh processing subunit, a calculation subunit, and an eighth processing subunit. The seventh processing subunit is used to perform preliminary feature selection on the target remote sensing image to obtain N image features, where N is an integer greater than 1; the calculation subunit is used to calculate the regression coefficient of each image feature in the N image features, where the regression coefficient of each image feature is used to characterize the degree of influence of the image feature on determining the health status information and growth information of the target tree; the eighth processing subunit is used to determine the target image feature from the N image features according to the regression coefficient.
[0088] Optionally, the processing unit 405 includes: a first determining subunit, a ninth processing subunit, and a tenth processing subunit. The first determining subunit is used to determine the value of the target tree according to the growth information and health status information of the target tree; the ninth processing subunit is used to approve the loan request if the value of the target tree is greater than or equal to the loan funds applied in the loan request; and the tenth processing subunit is used to generate risk warning information if the value of the target tree is less than the loan funds applied in the loan request.
[0089] Optionally, the tree-based loan request processing device further includes: a first acquisition unit, a second acquisition unit, a division unit, and a training unit. The first acquisition unit is used to acquire the growth and health status information of the reference tree of the same type as the target tree at Y historical moments, where Y is an integer greater than 1; the second acquisition unit is used to acquire Z historical images taken for the reference tree at Y historical moments, where each of the Y historical moments corresponds to at least one historical image, and Z is an integer greater than or equal to Y; the division unit is used to divide the Z historical images into X image sets according to the weather type of the area where the target tree is located at each historical moment, where the historical images in each image set correspond to the same weather type; the training unit is used to obtain X prediction models through training based on the X image sets, where the training data of each prediction model includes historical images in an image set and the growth and health status information of the reference tree when the historical images in the image set are taken.
[0090] Optionally, the statistical data is any one of a maximum value, a minimum value, a mean value, a median, and a mode.
[0091] According to another aspect of an embodiment of the present application, there is also provided an electronic device, comprising: a processor; and a memory for storing executable instructions of the processor; wherein the processor is configured to execute any one of the above-mentioned tree-based loan request processing methods by executing the executable instructions.
[0092] Figure 5 is a schematic diagram of an electronic device provided according to an embodiment of the present application, such as Figure 5 As shown, an embodiment of the present application provides an electronic device, which includes a processor, a memory, and a program stored in the memory and executable on the processor. When the processor executes the program, any one of the above-mentioned tree-based loan request processing methods is implemented.
[0093] The serial numbers of the above-mentioned embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.
[0094] In the above embodiments of the present application, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, please refer to the relevant description of other embodiments.
[0095] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the units can be a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.
[0096] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.
[0097] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.
[0098] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, disk or optical disk and other media that can store program codes.
[0099] The above is only a preferred implementation of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.
Claims
1. A tree-based loan request processing method, characterized in that: include: Acquire a target remote sensing image taken of a target tree under target weather conditions; Extracting target image features of the target tree in the target remote sensing image; Determine at least one target prediction model from X pre-trained prediction models according to the weather type of the target weather, wherein the X prediction models correspond to X weather types, and X is an integer greater than 1; The target image features are converted into a target vector by the target prediction model, and the growth potential information and health status information of the target tree are determined according to the target vector based on the prior knowledge learned during the model training process; The loan request using the target tree as collateral is processed based on the growth information and health status information of the target tree.
2. The tree-based loan request processing method according to claim 1, characterized in that: Determining at least one target prediction model from pre-trained X prediction models according to the weather type of the target weather, comprising: When the weather type of the target weather is the i-th weather type among the X weather types, the prediction model corresponding to the i-th weather type is used as the target prediction model, where i is a positive integer less than or equal to X.
3. The tree-based loan request processing method according to claim 1, characterized in that: Determining at least one target prediction model from pre-trained X prediction models according to the weather type of the target weather, comprising: When the weather type of the target weather is a mixed weather type composed of M weather types among the X weather types, the M prediction models corresponding to the M weather types are used as M target prediction models, where M is an integer less than or equal to X and greater than 1.
4. The tree-based loan request processing method according to claim 3, characterized in that: The target image features are converted into a target vector by the target prediction model, and the growth information and health status information of the target tree are determined according to the target vector based on the prior knowledge learned during the model training process, including: After taking the M prediction models corresponding to the M weather types as M target prediction models, converting the target image features into target vectors respectively through each of the M target prediction models; Determine the health score and growth score of the target tree according to the target vector using the prior knowledge learned by each target prediction model during the model training process, and obtain M health scores and M growth scores, wherein the M health scores are respectively determined by the M prediction models, and the M growth scores are respectively determined by the M prediction models; Obtain a target health score by calculating the statistical data of the M health scores, and characterize the health information of the target tree by the target health score; By calculating the statistical data of the M growth potential scores, a target growth potential score is obtained, and the growth potential information of the target tree is represented by the target growth potential score.
5. The tree-based loan request processing method according to claim 1, characterized in that: Extracting target image features of the target tree in the target remote sensing image includes: Performing preliminary feature selection on the target remote sensing image to obtain N image features, where N is an integer greater than 1; Calculating the regression coefficient of each of the N image features, wherein the regression coefficient of each image feature is used to characterize the degree of influence of the image feature on determining the health status information and growth potential information of the target tree; The target image feature is determined from the N image features according to the regression coefficient.
6. The tree-based loan request processing method according to claim 1, characterized in that: Processing a loan request with the target tree as collateral based on the growth information and health status information of the target tree includes: Determining the value of the target tree based on the growth information and health status information of the target tree; If the value of the target trees is greater than or equal to the loan amount applied for in the loan request, approve the loan request; In the case where the value of the target tree is less than the loan funds applied for in the loan request, risk warning information is generated.
7. The tree-based loan request processing method according to claim 1, characterized in that: The X prediction models are trained by the following steps, including: Obtaining growth information and health status information of reference trees of the same type as the target tree at Y historical moments, where Y is an integer greater than 1; Acquire Z historical images taken for the reference tree at the Y historical moments, wherein at least one historical image is taken corresponding to each of the Y historical moments, and Z is an integer greater than or equal to Y; Dividing the Z historical images into X image sets according to the weather type of the area where the target tree is located at each historical moment, wherein the historical images in each image set correspond to the same weather type; X prediction models are obtained by training according to the X image sets, wherein the training data of each prediction model includes a historical image in an image set and the growth potential information and health status information of the reference tree when the historical image in the image set is taken.
8. The tree-based loan request processing method according to claim 4, characterized in that: The statistical data is any one of the maximum value, minimum value, average value, median and mode.
9. A tree-based loan request processing device, characterized in that: include: An acquisition unit, used for acquiring a target remote sensing image taken of a target tree under target weather conditions; A feature extraction unit, used for extracting target image features of the target tree in the target remote sensing image; A first determining unit, configured to determine at least one target prediction model from X pre-trained prediction models according to the weather type of the target weather, wherein the X prediction models correspond to X weather types, and X is an integer greater than 1; A second determination unit is used to convert the target image features into a target vector through the target prediction model, and determine the growth information and health status information of the target tree according to the target vector based on the prior knowledge learned during the model training process; The processing unit is used to process the loan request using the target tree as collateral according to the growth information and health status information of the target tree.
10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, wherein the computer program is configured to execute the tree-based loan request processing method according to any one of claims 1 to 8 when executed.
11. An electronic device, characterized in that: The electronic device includes one or more processors; a storage device for storing one or more programs, which, when executed by the one or more processors, enables the one or more processors to implement the operation of the programs, wherein the programs are configured to execute the tree-based loan request processing method described in any one of claims 1 to 8 when running.