Wine-brewing sorghum yield prediction method and device, medium and product
By collecting and processing multispectral images using drones, the optimal remote sensing variables and texture features are screened out and input into the yield prediction model, which solves the time-consuming and labor-intensive problem of brewing sorghum yield prediction in traditional methods and achieves high-precision automated prediction.
Patent Information
- Application Number
- CN202510941100.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-07-08
- Publication Date
- 2025-09-26
AI Technical Summary
Traditional brewing sorghum yield prediction methods are time-consuming and labor-intensive, and difficult to expand to large-scale, timely and accurate predictions, which affects prediction accuracy.
Multispectral images were collected using drones, and through multispectral image processing and feature screening, multiple optimal remote sensing variables and texture features of brewing sorghum were determined, and then input into the trained yield prediction model for prediction.
The accuracy of brewing sorghum yield prediction has been improved, and automated, non-destructive, large-scale and timely prediction has been achieved.
Smart Images
Figure CN120707840A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field of yield prediction, and in particular to a method, device, medium and product for predicting the yield of brewing sorghum. Background Art
[0002] In recent years, the planting area and yield of brewing sorghum have continued to increase. During the planting process of brewing sorghum, it is very important to predict its yield.
[0003] Traditional yield forecasting methods are mainly based on spot checks of ground data. This method is time-consuming, labor-intensive, and destructive. It is difficult to expand to large-scale, timely, and accurate predictions of sorghum yields, which affects the accuracy of yield forecasts. Summary of the Invention
[0004] The purpose of this application is to provide a method, device, medium and product for predicting the yield of brewing sorghum to solve the problem of low accuracy in brewing sorghum yield prediction.
[0005] To achieve the above objectives, this application provides the following solutions:
[0006] In a first aspect, the present application provides a method for predicting brewing sorghum yield, comprising:
[0007] Acquire a target multispectral image; the target multispectral image is a multispectral image including a canopy of the brewing sorghum to be predicted at maturity;
[0008] Based on the target multispectral image, determining a multispectral orthophoto image of the brewing sorghum to be predicted after removing the background;
[0009] Based on the background-removed multispectral orthophoto of the brewing sorghum to be predicted, multiple optimal remote sensing variables and multiple optimal texture features of the brewing sorghum to be predicted are determined; the multiple optimal remote sensing variables are obtained by screening multiple initial remote sensing variables, and the multiple optimal texture features are obtained by screening multiple initial texture features;
[0010] Multiple optimal remote sensing variables and multiple optimal texture features of the brewing sorghum to be predicted are input into the yield prediction model to obtain the predicted value of the yield of the brewing sorghum to be predicted; the yield prediction model is obtained by screening the trained initial network, and the initial network includes: random forest network, support vector machine, partial least squares regression method and back propagation neural network.
[0011] In one embodiment, the target multispectral image is collected using a drone.
[0012] In one embodiment, determining a background-removed multispectral orthoimage of brewing sorghum to be predicted based on the target multispectral image includes:
[0013] Obtain the radiation correction parameters of the diffuse reflectance gray cloth when the UAV collects the target multispectral image;
[0014] Using the radiation correction parameters, the target multispectral image is subjected to two-dimensional multispectral reconstruction to obtain a multispectral orthophoto image of the brewing sorghum to be predicted;
[0015] The multispectral orthophoto image of the brewing sorghum to be predicted is subjected to mask processing to obtain the multispectral orthophoto image of the brewing sorghum to be predicted with the background removed.
[0016] In one embodiment, based on a multispectral orthoimage of the brewing sorghum to be predicted after removing the background, a plurality of optimal remote sensing variables and a plurality of optimal texture features of the brewing sorghum to be predicted are determined, including:
[0017] Determine the canopy spectral values of multiple bands of the brewing sorghum to be predicted based on the multispectral orthophoto image after removing the background of the brewing sorghum to be predicted;
[0018] According to the canopy spectral values of multiple bands of the brewing sorghum to be predicted, multiple optimal remote sensing variables of the brewing sorghum to be predicted are determined;
[0019] Based on the multispectral orthophoto image of the brewing sorghum to be predicted with the background removed, multiple optimal texture features of the brewing sorghum to be predicted are determined.
[0020] In one embodiment, the plurality of wavelength bands include a green wavelength band, a red wavelength band, a red-edge wavelength band, and a near-infrared wavelength band.
[0021] In one embodiment, the plurality of optimal remote sensing variables include: canopy spectral values in the red light band, normalized difference vegetation index, ratio vegetation index, improved simple ratio vegetation index, and nitrogen reflectance index;
[0022] Multiple optimal texture features include: the average grayscale value of the green light band, the average grayscale value of the red light band, the correlation of the green light band, the correlation of the red edge band and the correlation of the near-infrared band of each area of interest in the multispectral orthophoto of the brewing sorghum to be predicted after removing the background.
[0023] In one embodiment, the process of determining the yield prediction model includes:
[0024] Acquire an initial data set; the initial data set includes: a plurality of optimal remote sensing variables, a plurality of optimal texture features, and measured values of yield of a plurality of sample brewing sorghum;
[0025] Dividing the initial data set into a training set and a test set according to a preset ratio;
[0026] Using the training set to train the initial networks respectively to obtain multiple trained initial networks;
[0027] Using the test set, determining the coefficient of determination of each trained initial network;
[0028] The trained initial network with the largest determination coefficient is determined as the yield prediction model.
[0029] In a second aspect, the present application provides a computer device comprising: 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 above-mentioned brewing sorghum yield prediction method.
[0030] In a third aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-mentioned method for predicting brewing sorghum yield.
[0031] In a fourth aspect, the present application provides a computer program product, comprising a computer program, which, when executed by a processor, implements the above-mentioned method for predicting brewing sorghum yield.
[0032] According to the specific embodiments provided in this application, this application discloses the following technical effects:
[0033] The present application discloses a method, device, medium and product for predicting the yield of brewing sorghum. First, a target multispectral image is obtained; the target multispectral image is a multispectral image of the canopy of the brewing sorghum to be predicted at maturity; then, based on the target multispectral image, a multispectral orthophoto image of the brewing sorghum to be predicted with the background removed is determined; subsequently, based on the multispectral orthophoto image of the brewing sorghum to be predicted with the background removed, multiple optimal remote sensing variables and multiple optimal texture features of the brewing sorghum to be predicted are determined; the multiple optimal remote sensing variables are obtained by screening multiple initial remote sensing variables, and the multiple optimal texture features are obtained by screening multiple initial texture features; finally, the multiple optimal remote sensing variables and multiple optimal texture features of the brewing sorghum to be predicted are input into a yield prediction model to obtain a predicted value of the yield of the brewing sorghum to be predicted; the yield prediction model is obtained by screening a trained initial network, and the initial network includes: a random forest network, a support vector machine, a partial least squares regression method and a back propagation neural network. This application inputs the optimal remote sensing variables obtained by screening the initial remote sensing variables and the optimal texture features obtained by screening multiple initial texture features into the yield prediction model obtained by screening the trained initial network, and automatically determines the predicted value of the yield of the brewing sorghum to be predicted. Compared with the traditional yield prediction based on spot checks of ground data, the accuracy of brewing sorghum yield prediction is improved. BRIEF DESCRIPTION OF THE DRAWINGS
[0034] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.
[0035] Figure 1 This is an application environment diagram of a method for predicting brewing sorghum yield in one embodiment of the present application;
[0036] Figure 2 A schematic diagram of a process flow for predicting brewing sorghum yield provided in one embodiment of the present application;
[0037] Figure 3 This is the result of ranking the importance of remote sensing variables of Jinnuoliang by RFE;
[0038] Figure 4 This is the result of the decision tree ranking the importance of remote sensing variables for Jinnuoliang;
[0039] Figure 5 This is the result of ranking the importance of remote sensing variables of RFE to red tassel;
[0040] Figure 6 This is the result of the decision tree ranking the importance of remote sensing variables of red tassel;
[0041] Figure 7 This is the result of RFE ranking the importance of texture features of Jinnuoliang;
[0042] Figure 8 This is the result of the decision tree ranking the importance of texture features of Jinnuoliang;
[0043] Figure 9 This is the result of RFE ranking the importance of texture features of red tassels;
[0044] Figure 10 This is the result of the decision tree ranking the importance of the texture features of red tassels;
[0045] Figure 11 It is a curve diagram showing the relationship between the measured value and the predicted value;
[0046] Figure 12 A schematic diagram of the structure of a computer device provided in one embodiment of the present application. DETAILED DESCRIPTION
[0047] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.
[0048] The purpose of this application is to provide a method, device, medium and product for predicting the yield of brewing sorghum, aiming to improve the accuracy of brewing sorghum yield prediction.
[0049] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.
[0050] The brewing sorghum yield prediction method provided in the embodiment of the present application can be applied to Figure 1 In the application environment shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be set up separately, integrated on the server 104, or placed on the cloud or other servers. The terminal 102 can send a multispectral image including the canopy of the brewing sorghum to be predicted at the maturity stage to the server 104. After the server 104 receives the multispectral image including the canopy of the brewing sorghum to be predicted at the maturity stage, for the multispectral image including the canopy of the brewing sorghum to be predicted at the maturity stage, the server 104 determines the multispectral orthophoto image of the brewing sorghum to be predicted after removing the background based on the target multispectral image; based on the multispectral orthophoto image of the brewing sorghum to be predicted after removing the background, determine multiple optimal remote sensing variables and multiple optimal texture features of the brewing sorghum to be predicted; input the multiple optimal remote sensing variables and multiple optimal texture features of the brewing sorghum to be predicted into the yield prediction model to obtain the predicted value of the yield of the brewing sorghum to be predicted. The server 104 can feed back the obtained predicted value of the yield of the brewing sorghum to be predicted to the terminal 102. In addition, in some embodiments, the brewing sorghum yield prediction method can also be implemented independently by the server 104 or the terminal 102. For example, the terminal 102 can directly predict the brewing sorghum yield based on the multispectral image of the canopy of the brewing sorghum to be predicted at the maturity stage, or the server 104 can obtain the multispectral image of the canopy of the brewing sorghum to be predicted at the maturity stage from the data storage system, and predict the brewing sorghum yield based on the multispectral image of the canopy of the brewing sorghum to be predicted at the maturity stage.
[0051] In an exemplary embodiment, Figure 2 As shown, a method for predicting brewing sorghum yield is provided, comprising:
[0052] Step 1: Acquire a target multispectral image; the target multispectral image is a multispectral image of the canopy of the brewing sorghum to be predicted at maturity.
[0053] Specifically, during the mature stage, the grains of brewing sorghum are bright and lustrous. They are firm and resistant to pinching when pressed with fingernails. At this stage, the grains are relatively dry and produce a crisp sound when bitten. At maturity, the ears droop, forming a natural curve. The rachis are tough and resistant to breaking. The shape of the ear is essentially fixed, with no noticeable growth changes. The husks dry out and fall off easily. The lower leaves of the plant turn yellow and dry, spreading upward. The stems become dry, hard, and lighter in color, with a significantly reduced moisture content.
[0054] As an optional implementation, the target multispectral image is collected using a drone.
[0055] Specifically, the drone is the DJI Mavic 3 Multispectral.
[0056] Step 2: Based on the target multispectral image, determine the multispectral orthophoto image of the brewing sorghum to be predicted after removing the background.
[0057] As an optional implementation, step 2 includes:
[0058] Step 21: Obtain the radiation correction parameters of the diffuse gray cloth when the UAV collects the target multispectral image.
[0059] Step 22: Use the radiation correction parameters to perform two-dimensional multispectral reconstruction on the target multispectral image to obtain a multispectral orthophoto image of the brewing sorghum to be predicted.
[0060] Specifically, the pixel value at a pixel point in the multispectral orthoimage corresponds to the reflectance of each band.
[0061] Step 23: Masking is performed on the multispectral orthophoto image of the wine-making sorghum to be predicted, so as to obtain a multispectral orthophoto image of the wine-making sorghum to be predicted with the background removed.
[0062] Specifically, the multispectral orthophoto image to be predicted for brewing sorghum is subjected to mask processing, so as to remove the background portion of the multispectral orthophoto image.
[0063] Step 3: Based on the multispectral orthophoto image of the brewing sorghum to be predicted after removing the background, determine multiple optimal remote sensing variables and multiple optimal texture features of the brewing sorghum to be predicted.
[0064] The multiple optimal remote sensing variables are obtained by screening the multiple initial remote sensing variables, and the multiple optimal texture features are obtained by screening the multiple initial texture features.
[0065] As an optional implementation, step 3 includes:
[0066] Step 31: Determine the canopy spectral values of multiple bands of the brewing sorghum to be predicted based on the multispectral orthophoto image of the brewing sorghum to be predicted after removing the background.
[0067] Specifically, the mean of the reflectance of any band at all pixels in the multispectral orthoimage after background removal is taken as the canopy spectral value of the band.
[0068] As an optional implementation, the multiple bands include: a green light band, a red light band, a red edge band and a near infrared band.
[0069] Specifically, the wavelength of the green light band is 560 nm, the wavelength of the red light band is 650 nm, the wavelength of the red edge band is 730 nm, and the wavelength of the near infrared band is 860 nm.
[0070] Step 32: Determine multiple optimal remote sensing variables for the brewing sorghum to be predicted based on the canopy spectral values of multiple bands of the brewing sorghum to be predicted.
[0071] Specifically, the screening process of multiple optimal remote sensing variables includes:
[0072] S11: Obtain canopy spectral values of multiple bands of the sample brewing sorghum, and determine multiple initial remote sensing variables based on the canopy spectral values of multiple bands of the sample sorghum. The initial remote sensing variables are shown in Table 1.
[0073] Table 1 Initial remote sensing variables
[0074]
[0075] Among them, SQRT means finding the square root.
[0076] S12: The importance ranking and feature screening of 18 initial remote sensing variables were performed using the recursive feature elimination (RFE) method based on cross-validation. The initial remote sensing variables with importance values greater than 0.05 were selected as the optimal remote sensing variables after RFE screening. The importance ranking results of RFE for the initial texture features of different varieties of sorghum (golden glutinous sorghum and red tassel) were as follows: Figure 3 and Figure 5 shown. Figure 3 and Figure 5 The middle vertical axis represents the importance value. The optimal remote sensing variables after RFE screening include: R Red , NDVI, MSR, RVI, NRI.
[0077] S13: Use decision tree to sort the importance and select features of 18 initial remote sensing variables, and select the initial remote sensing variables with importance values greater than 0.05 as the optimal remote sensing variables after decision tree screening. The importance sorting results of the decision tree for the initial texture features of different varieties of sorghum (golden glutinous sorghum and red tassel) are as follows: Figure 4 and Figure 6 shown. Figure 4 and Figure 6 The middle vertical axis represents the importance value. The optimal remote sensing variables after the decision tree screening include: MSR, RVI, R Red , NDVI, SRCI, GNDVI, NRI, CIgreen.
[0078] S14: The intersection of the optimal remote sensing variables after RFE screening and the optimal remote sensing variables after decision tree screening is determined as multiple optimal remote sensing variables, namely R Red , NDVI, MSR, RVI, NRI.
[0079] Step 33: Determine multiple optimal texture features of the brewing sorghum to be predicted based on the multispectral orthophoto image of the brewing sorghum to be predicted after removing the background.
[0080] Specifically, the screening process of multiple optimal texture features includes:
[0081] S21: Obtain a multispectral orthophoto image of the sample brewing sorghum after removing the background, and determine multiple initial texture features based on the multispectral orthophoto image of the brewing sorghum to be predicted after removing the background. The initial texture features are shown in Table 2.
[0082] Table 2 Initial texture feature table
[0083]
[0084] Wherein, X_mean represents the average grayscale value of a certain band of a certain region of interest in the multispectral orthoimage after background removal (when X is G, it is the green band; when X is R, it is the red band; when X is Rededge, it is the red edge band; when X is NIR, it is the near infrared band); X_var represents the variance of a certain band of a certain region of interest in the multispectral orthoimage after background removal; X_hom represents the homogeneity of a certain band of a certain region of interest in the multispectral orthoimage after background removal; X_Con represents the contrast of a certain band of a certain region of interest in the multispectral orthoimage after background removal; X_dis represents the difference of a certain band of a certain region of interest in the multispectral orthoimage after background removal; X_ent represents the entropy of a certain band of a certain region of interest in the multispectral orthoimage after background removal; X_sm represents the second-order moment of a certain band of a certain region of interest in the multispectral orthoimage after background removal; X_cor represents the correlation of a certain band of a certain region of interest in the multispectral orthoimage after background removal.
[0085] S22: RFE was used to sort the importance of 32 initial texture features and screen the features. The initial texture features with importance values greater than 0.03 were selected as the optimal texture features after RFE screening. The importance sorting results of RFE on the initial texture features of different varieties of sorghum (golden glutinous sorghum and red tassel) were as follows: Figure 7 and Figure 9 shown. Figure 7 and Figure 9 The vertical axis represents the importance value. The optimal texture features after RFE screening include: R_mean, G_mean, Rededge_cor, NIR_cor, G_cor, and Rededge_mean.
[0086] S23: Use decision tree to sort the importance of 32 initial texture features and select features. The initial texture features with importance values greater than 0.03 are taken as the optimal texture features after the decision tree screening. The importance sorting results of the decision tree for the initial texture features of different varieties of sorghum (golden glutinous sorghum and red tassel) are as follows: Figure 8 and Figure 10 shown. Figure 8 and Figure 10 The vertical axis represents the importance value. The optimal texture features after the decision tree screening include: G_mean, G_cor, R_mean, Rededge_cor, NIR_mean, NIR_cor and Rededge_sm.
[0087] S24: Determine the intersection of the optimal texture features after RFE screening and the optimal texture features after decision tree screening as multiple optimal texture features, namely R_mean, G_mean, Rededge_cor, NIR_cor, and G_cor.
[0088] As an optional implementation, the multiple optimal remote sensing variables include: canopy spectral values in the red light band, normalized difference vegetation index, ratio vegetation index, improved simple ratio vegetation index and nitrogen reflectance index.
[0089] Multiple optimal texture features include: the average grayscale value of the green light band, the average grayscale value of the red light band, the correlation of the green light band, the correlation of the red edge band and the correlation of the near-infrared band of each area of interest in the multispectral orthophoto of the brewing sorghum to be predicted after removing the background.
[0090] Step 4: Input multiple optimal remote sensing variables and multiple optimal texture features of the brewing sorghum to be predicted into the yield prediction model to obtain the predicted value of the yield of the brewing sorghum to be predicted.
[0091] Among them, the yield prediction model is obtained by screening the trained initial networks, which include: Random Forest Network (RF), Support Vector Machine (SVM), Partial Least Squares Regression (PLSR) and Backpropagation Neural Network (BPNN).
[0092] As an optional implementation, in step 4, the process of determining the yield prediction model includes:
[0093] Step 41: Obtain an initial data set; the initial data set includes: multiple optimal remote sensing variables, multiple optimal texture features and measured values of yield of multiple samples of brewing sorghum.
[0094] Specifically, the harvested sample brewing sorghum grains were threshed and air-dried, and then their weight and moisture content were measured. The yields obtained at different moisture contents were then normalized and converted to the calculated result when the moisture content was 14%, and the result was used as the actual measured yield value.
[0095] Step 42: Divide the initial data set into a training set and a test set according to a preset ratio.
[0096] Step 43: Use the training set to train the initial networks respectively to obtain multiple trained initial networks.
[0097] Step 44: Using the test set, determine the coefficient of determination of each trained initial network.
[0098] Specifically, in determining the coefficient of determination R 2 When , the root mean square error (RMSE) of each trained initial network can also be determined, R 2 The higher the value, the lower the RMSE, which indicates that the model has higher accuracy. The coefficient of determination and root mean square error of each trained initial network are shown in Table 3.
[0099] Table 3 Determination coefficient and root mean square error
[0100]
[0101]
[0102] It can be seen that the R of each model during sorghum maturity is 2 The range is 0.56~0.67, and the R 2 The highest, and its corresponding RMSE is also the lowest, which is 204.08 kg·ha -1 , indicating that this application can be applied to the non-destructive estimation of general brewing sorghum yield.
[0103] The present invention also uses a validation set to verify the method of the present invention. The relationship curve between the measured value and the predicted value of the yield of the validation set is shown in the figure below. Figure 11 shown.
[0104] Step 45: The trained initial network with the largest determination coefficient is determined as the yield prediction model.
[0105] In an exemplary embodiment, a computer device is provided, comprising: 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 a method for predicting brewing sorghum yield.
[0106] In an exemplary embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, a method for predicting the yield of brewing sorghum is implemented.
[0107] In an exemplary embodiment, a computer program product is provided, comprising a computer program that implements a method for predicting brewing sorghum yield when executed by a processor.
[0108] In an exemplary embodiment, a computer device is provided. The computer device may be a server or a terminal. The internal structure diagram thereof may be as follows: Figure 12As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, abbreviated as I / O) and a communication interface. The processor, memory and input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for predicting the yield of brewing sorghum is implemented.
[0109] Those skilled in the art will understand that Figure 12 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.
[0110] Those skilled in the art will understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, database or other media used in the embodiments provided in this application may include at least one of non-volatile and volatile memory. Non-volatile memory may include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory may include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM may be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).
[0111] The databases involved in the various embodiments provided herein may include at least one of a relational database and a non-relational database. Non-relational databases may include, but are not limited to, distributed databases based on blockchains. The processors involved in the various embodiments provided herein may include, but are not limited to, general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic units, data processing logic units based on quantum computing, and the like.
[0112] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.
[0113] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.
[0114] This document uses specific examples to illustrate the principles and implementation methods of this application. The description of the above examples is only intended to help understand the method and core concept of this application. At the same time, for those skilled in the art, based on the concept of this application, there may be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.
Claims
1. A method for predicting brewing sorghum yield, characterized in that: The brewing sorghum yield prediction method comprises: Acquire a target multispectral image; the target multispectral image is a multispectral image including a canopy of the brewing sorghum to be predicted at maturity; Based on the target multispectral image, determining a multispectral orthophoto image of the brewing sorghum to be predicted after removing the background; Based on the background-removed multispectral orthophoto of the brewing sorghum to be predicted, multiple optimal remote sensing variables and multiple optimal texture features of the brewing sorghum to be predicted are determined; the multiple optimal remote sensing variables are obtained by screening multiple initial remote sensing variables, and the multiple optimal texture features are obtained by screening multiple initial texture features; Multiple optimal remote sensing variables and multiple optimal texture features of the brewing sorghum to be predicted are input into the yield prediction model to obtain the predicted value of the yield of the brewing sorghum to be predicted; the yield prediction model is obtained by screening the trained initial network, and the initial network includes: random forest network, support vector machine, partial least squares regression method and back propagation neural network.
2. The method for predicting brewing sorghum yield according to claim 1, wherein: The target multispectral image is collected by using a drone.
3. The method for predicting brewing sorghum yield according to claim 2, wherein: Determining a background-removed multispectral orthophoto of brewing sorghum to be predicted based on the target multispectral image includes: Obtain the radiation correction parameters of the diffuse reflectance gray cloth when the UAV collects the target multispectral image; Using the radiation correction parameters, the target multispectral image is subjected to two-dimensional multispectral reconstruction to obtain a multispectral orthophoto image of the brewing sorghum to be predicted; The multispectral orthophoto image of the brewing sorghum to be predicted is subjected to mask processing to obtain the multispectral orthophoto image of the brewing sorghum to be predicted with the background removed.
4. The method for predicting brewing sorghum yield according to claim 1, wherein: Based on the background-removed multispectral orthophoto of the brewing sorghum to be predicted, multiple optimal remote sensing variables and multiple optimal texture features of the brewing sorghum to be predicted are determined, including: Determine the canopy spectral values of multiple bands of the brewing sorghum to be predicted based on the multispectral orthophoto image after removing the background of the brewing sorghum to be predicted; According to the canopy spectral values of multiple bands of the brewing sorghum to be predicted, multiple optimal remote sensing variables of the brewing sorghum to be predicted are determined; Based on the multispectral orthophoto image of the brewing sorghum to be predicted with the background removed, multiple optimal texture features of the brewing sorghum to be predicted are determined.
5. The method for predicting brewing sorghum yield according to claim 4, characterized in that: Multiple bands include: green light band, red light band, red edge band and near infrared band.
6. The method for predicting brewing sorghum yield according to claim 5, characterized in that: Several optimal remote sensing variables include: canopy spectral values in the red band, normalized difference vegetation index, ratio vegetation index, improved simple ratio vegetation index, and nitrogen reflectance index; Multiple optimal texture features include: the average grayscale value of the green light band, the average grayscale value of the red light band, the correlation of the green light band, the correlation of the red edge band and the correlation of the near-infrared band of each area of interest in the multispectral orthophoto of the brewing sorghum to be predicted after removing the background.
7. The method for predicting brewing sorghum yield according to claim 1, wherein: The process of determining the yield prediction model includes: Acquire an initial data set; the initial data set includes: a plurality of optimal remote sensing variables, a plurality of optimal texture features, and measured values of yield of a plurality of sample brewing sorghum; Dividing the initial data set into a training set and a test set according to a preset ratio; Using the training set to train the initial networks respectively to obtain multiple trained initial networks; Using the test set, determining the coefficient of determination of each trained initial network; The trained initial network with the largest determination coefficient is determined as the yield prediction model.
8. A computer device comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the computer program to implement the brewing sorghum yield prediction method according to any one of claims 1 to 7.
9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for predicting the yield of brewing sorghum described in any one of claims 1 to 7 is implemented.
10. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the method for predicting the yield of brewing sorghum described in any one of claims 1 to 7 is implemented.