Insect body axis ratio inversion method, device and medium based on RCS characteristics

Abstract

The present application provides an insect body axis ratio inversion method, device and medium based on RCS characteristics, which belongs to the field of insect radar technology. The method is based on the analysis results of the preset insect collinear polarization pattern, determines the correlation between the characteristic parameters of each pattern and the insect body axis ratio, and determines multiple insect body axis ratio estimators based on the correlation relationship; wherein the correlation relationship includes at least correlation with the insect body length and correlation with the insect body width; the insect body axis ratio estimator is used to characterize the quantitative relationship between the insect polarization characteristics and the insect body axis ratio; according to each insect body axis ratio estimator and the preset XGBoost regression algorithm, a pre-trained body axis ratio inversion model is constructed; wherein the body axis ratio inversion model establishes a mapping relationship between each insect body axis ratio estimator and the insect body axis ratio; the body axis ratio inversion model is deployed to the user terminal, so that the user terminal performs insect body axis ratio inversion based on the input insect collinear polarization pattern and the body axis ratio inversion model.

Description

Technical Field

[0001] The present application relates to the field of insect radar technology, and in particular to an insect body axis ratio inversion method, device, and medium based on RCS characteristics. Background Art

[0002] Insect radar is a tool specifically designed for monitoring and studying migratory insects. It enables all-weather, all-day monitoring without interfering with insect migration, and is therefore widely used in insect research. Insect echoes measured by insect radar can be used to estimate insect biological and behavioral characteristics, such as body length, weight, wingbeat frequency, horizontal flight speed, and direction. These parameters enable species identification and trajectory analysis of migratory insects, which is crucial for developing effective aerial monitoring, interception, and control systems for migratory pests.

[0003] The insect body axis ratio (i.e., length-to-width ratio) is an important morphological feature for species identification. Traditional methods have established a mapping relationship between RCS estimators and morphological features by searching for radar cross-section (RCS) estimators associated with specific insect morphological features (such as weight and length), thereby enabling the inversion of these features. However, no RCS estimator related to the body axis ratio has been found, and traditional methods for measuring insect body axis ratio rely primarily on manual observation and physical measurement. This method is not only inefficient but also extremely difficult to measure for tiny insects or those in complex environments, and accuracy cannot be guaranteed. Accurately and efficiently inverting the insect body axis ratio remains an unsolved problem.

[0004] Therefore, in order to further improve the accuracy of migratory insect species identification, there is an urgent need for a technical solution that can accurately invert the body axis ratio of migratory insects. Summary of the Invention

[0005] The embodiments of the present application provide a method, device and medium for inverting the insect body axis ratio based on RCS features, which are used to solve the current technical problem that it is difficult to achieve efficient and accurate inversion of the insect body axis ratio.

[0006] On the one hand, an embodiment of the present application provides an insect body axis ratio inversion method based on RCS features, the method comprising:

[0007] Based on the analysis results of preset insect collinear polarization patterns, a correlation relationship between characteristic parameters of each pattern and the insect's body axis ratio is determined, and multiple insect body axis ratio estimators are determined based on the correlation relationship; wherein the correlation relationship includes at least a correlation with the insect body length in the insect body axis ratio and a correlation with the insect body width in the insect body axis ratio determined based on the co-polarization radar scattering cross section (RCS); the insect body axis ratio estimator is used to characterize the quantitative relationship between the insect polarization characteristics and the insect body axis ratio;

[0008] According to each of the insect body axis ratio estimators and a preset XGBoost regression algorithm, a pre-trained body axis ratio inversion model is constructed; wherein the body axis ratio inversion model establishes a mapping relationship between each of the insect body axis ratio estimators and the insect body axis ratio;

[0009] The body axis ratio inversion model is deployed to a user terminal, so that the user terminal performs insect body axis ratio inversion based on the input insect collinear polarization pattern and the body axis ratio inversion model.

[0010] In one implementation of the present application, based on the analysis results of the preset insect collinear polarization patterns, before determining the correlation between the characteristic parameters of each pattern and the insect body axis ratio, the method further includes:

[0011] Obtaining co-polarization radar cross sections (RCS) corresponding to different polarization directions of a pre-observed insect; wherein the co-polarization RCS is obtained by observing the pre-observed insect using a polarization insect radar in a linear polarization mode;

[0012] The preset insect collinear polarization pattern is determined according to the change of the co-polarization RCS in the polarization direction; wherein the preset insect collinear polarization pattern is expressed as:

[0013]

[0014] in, represents the insect collinear polarization pattern, is the characteristic parameter of the pattern that is independent of the polarization direction; and is a pattern characteristic parameter related to the shape of the insect's collinear polarization pattern; Indicates the polarization direction, and its value range is ; Indicates the direction of the insect's body axis.

[0015] In one implementation of the present application, based on the analysis results of the preset insect collinear polarization pattern, the correlation relationship between the characteristic parameters of each pattern and the insect body axis ratio is determined, specifically including:

[0016] Determine the preset insect collinear polarization pattern Item and When the items respectively reach their maximum values, the angle between the polarization direction and the insect body axis direction;

[0017] According to the angle side directions corresponding to the angles and the preset insect body length direction, respectively determining the body length direction and body width direction corresponding to the polarization directions;

[0018] According to the Item and The corresponding relationship between the item and the body length direction and the body width direction is established. and stated The correlation between them and the body axis ratio of the insect respectively.

[0019] In one implementation of the present application, determining a plurality of insect body axis ratio estimators based on the association relationship specifically includes:

[0020] According to the association relationship, when the angle is 0 or When As the RCS in the body length direction;

[0021] According to the association relationship, when the angle is or When As the RCS in the body width direction;

[0022] According to the association relationship, the ratio of the body length direction RCS to the body width direction RCS As an indicator of RCS body axis ratio;

[0023] The 、 , the RCS in the body length direction, the RCS in the body width direction and the RCS body axis ratio index are respectively used as the insect body axis ratio estimator.

[0024] In one implementation of the present application, a pre-trained axial ratio inversion model is constructed based on each of the insect axial ratio estimators and the preset XGBoost regression algorithm, specifically including:

[0025] Constructing a plurality of multidimensional feature vector samples based on each of the insect body axis ratio estimators and a plurality of insect collinear polarization pattern samples; wherein the multidimensional feature vector samples include a multidimensional feature vector composed of each of the insect body axis ratio estimators corresponding to each of the insect collinear polarization pattern samples and an insect body axis ratio label;

[0026] Each of the multidimensional feature vector samples is input into the preset XGBoost regression algorithm to train the preset XGBoost regression algorithm until a training end condition is met, thereby obtaining the body axis ratio inversion model.

[0027] In one implementation of the present application, the preset XGBoost regression algorithm is trained until the training end condition is met to obtain the body axis ratio inversion model, specifically including:

[0028] Iteratively training the preset XGBoost regression algorithm through each of the multidimensional feature vector samples in the preset training set;

[0029] Inputting each of the multidimensional feature vector samples in the preset test set into the preset XGBoost regression algorithm after iterative training to calculate the corresponding average relative error according to the output result;

[0030] When the average relative error is less than a preset threshold, the preset XGBoost regression algorithm after iterative training is determined as the body axis ratio inversion model.

[0031] In one implementation of the present application, after deploying the body axis ratio inversion model to the user terminal, the method further includes:

[0032] Obtaining each of the insect body axis ratio estimators corresponding to the input insect collinear polarization pattern, and constructing the corresponding multidimensional feature vector;

[0033] The multidimensional feature vector is sent to the user terminal, so that the user terminal performs insect body axis ratio inversion.

[0034] In one implementation of the present application, the user terminal performs insect body axis ratio inversion based on the input insect collinear polarization pattern and the body axis ratio inversion model, specifically including:

[0035] The user terminal determines a plurality of corresponding insect body axis ratio estimators according to the input insect collinear polarization pattern, and constructs the multidimensional feature vector;

[0036] The user terminal inputs the multi-dimensional feature vector into the pre-deployed body axis ratio inversion model to perform insect body axis ratio inversion.

[0037] On the other hand, an embodiment of the present application further provides an insect body axis ratio inversion device based on RCS features, the device comprising:

[0038] At least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the insect body axis ratio inversion method based on RCS features as described above.

[0039] On the other hand, an embodiment of the present application further provides a non-volatile computer storage medium storing computer executable instructions, wherein the computer executable instructions can execute the insect body axis ratio inversion method based on RCS features as described above.

[0040] Compared with the prior art, this application has the following significant effects:

[0041] (1) Through the above scheme, this application deeply analyzes the preset insect collinear polarization patterns, determines the correlation between the characteristic parameters of each pattern and the insect body axis ratio, and then determines multiple insect body axis ratio estimators. These estimators can accurately characterize the quantitative relationship between the insect polarization characteristics and the insect body axis ratio, providing an accurate data basis for the subsequent construction of the body axis ratio inversion model, thereby greatly improving the accuracy of the insect body axis ratio inversion. The body axis ratio inversion model constructed based on the preset XGBoost regression algorithm can quickly process the input insect polarization data. Compared with traditional methods, it greatly improves the efficiency of insect body axis ratio inversion, making it possible to perform body axis ratio analysis on a large number of insect samples in a short time.

[0042] (2) The body axis ratio inversion model is deployed to the user terminal. The user only needs to input the insect collinear polarization pattern in the terminal to perform insect body axis ratio inversion based on the model. This design gets rid of the dependence of traditional measurement methods on complex equipment and professional environments, allowing insect researchers to conveniently analyze insect body axis ratios in a variety of scenarios such as the field and the laboratory, significantly enhancing the ease of application of the technology. This application effectively solves the technical problem of the difficulty in achieving efficient and accurate inversion of insect body axis ratios, providing a more accurate, efficient and convenient research method for insect research. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] 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:

[0044] Figure 1 A schematic flow chart of an insect body axis ratio inversion method based on RCS features in an embodiment of the present application;

[0045] Figure 2Schematic diagram of a simulated collinear polarization pattern in an insect body axis ratio inversion method based on RCS characteristics in an embodiment of the present application;

[0046] Figure 3 Schematic diagram showing the comparison between the estimated body axis ratio and the true body axis ratio in an insect body axis ratio inversion method based on RCS features in an embodiment of the present application;

[0047] Figure 4 This is a structural schematic diagram of an insect body axis ratio inversion device based on RCS characteristics in an embodiment of the present application. DETAILED DESCRIPTION

[0048] To make the purpose, technical solutions, and advantages of this application more clear, the technical solutions of this application will be clearly and completely described below in conjunction with the specific embodiments of this application and the corresponding drawings. Obviously, the embodiments described are only part of the embodiments of this application, not all of them. 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.

[0049] In the field of insect research, accurately determining insect body axis ratios is crucial for gaining a deeper understanding of their ecological habits, behavioral patterns, and evolutionary characteristics. Traditional methods for measuring insect body axis ratios rely primarily on manual observation and physical measurement, which is not only inefficient but also extremely difficult and inaccurate for tiny insects or those in complex environments.

[0050] Based on this, the embodiments of the present application provide an insect body axis ratio inversion method, device and medium based on RCS characteristics, which are used to solve the current technical problem of difficulty in efficiently and accurately inverting the insect body axis ratio.

[0051] The following describes in detail various embodiments of the present application with reference to the accompanying drawings.

[0052] The embodiment of the present application provides an insect body axis ratio inversion method based on RCS characteristics, such as Figure 1 As shown, the method may include steps S101-S103:

[0053] S101, the server determines the correlation between characteristic parameters of each pattern and the insect body axis ratio based on the analysis results of the preset insect collinear polarization pattern, and determines multiple insect body axis ratio estimators based on the correlation.

[0054] The correlation relationship includes at least determining the correlation between the insect's body length in the insect's body axis ratio and the insect's body width in the insect's body axis ratio based on the co-polarization radar cross section (RCS). The insect's body axis ratio estimator is used to characterize the quantitative relationship between the insect's polarization characteristics and the insect's body axis ratio.

[0055] It should be noted that the server, as the executor of the insect body axis ratio inversion method based on RCS features, is only an example. The executor is not limited to the server, and this application does not make any specific restrictions on this.

[0056] In the embodiment of the present application, based on the analysis results of the preset insect collinear polarization pattern, before determining the correlation between the characteristic parameters of each pattern and the insect body axis ratio, the following steps are further included:

[0057] Obtain the co-polarization radar cross section (RCS) corresponding to the previously observed insect in different polarization directions. The co-polarization RCS is obtained by observing the previously observed insect using a polarization insect radar in linear polarization mode. Based on the variation of the co-polarization RCS across polarization directions, determine the collinear polarization pattern of the preset insect. The collinear polarization pattern of the preset insect is expressed as:

[0058]

[0059] in, represents the insect collinear polarization pattern, It is a characteristic parameter of the directional pattern that is independent of the polarization direction. and is a characteristic parameter of the pattern related to the shape of the insect's collinear polarization pattern and is a dimensionless parameter. Indicates the polarization direction, and its value range is . Indicates the direction of the insect's body axis.

[0060] That is to say, this application analyzes the relationship between the polarization characteristics and the insect's body axis ratio in the insect's collinear polarization pattern, and calculates the insect's body axis ratio estimator. Specifically, this application uses a polarization insect radar to observe with linear polarization, and can obtain the co-polarization RCS of the insect in different polarization directions. Under the assumption that the insect is symmetric, this application expresses the co-polarization RCS as above .

[0061] In one embodiment of the present application, the above-mentioned determination of the correlation between the characteristic parameters of each pattern and the insect body axis ratio based on the analysis results of the preset insect collinear polarization pattern specifically includes:

[0062] Determine the preset insect collinear polarization pattern Item and When the terms are maximized, the angle between the polarization direction and the insect body axis is determined. According to the angle side direction corresponding to each angle and the preset insect body length direction, the body length direction and body width direction corresponding to the polarization direction are determined respectively. Xiang He The corresponding relationship between the item and the body length direction and body width direction is established. and Their respective correlations with the insect body axis ratios.

[0063] In other words, the present application can calculate the collinear polarization pattern. Item and The angle between the polarization direction and the insect body axis when the terms reach their maximum values. The period of the term is , its maximum value appears at , At this time, the polarization direction is aligned with the preset insect body length direction, which is the body length direction, that is, Related to the elongation of the collinear polarization pattern along the body length. The period is , its maximum value appears at , , , , since the above-mentioned , Indicates the direction of body length, , It means that the polarization direction is perpendicular to the body axis, and the polarization direction is the body width direction, that is, It is correlated with the collinear polarization pattern along both the length and width of the insect body, which can be more generally explained as being correlated with the cross shape of the collinear polarization pattern.

[0064] Figure 2 The normalized collinear polarization pattern of four simulated insects is shown in Figure 1. The body length of the four insects is 15 mm, and the preset body length direction of the insects is 0-180°. Figure 2 From (a) to (d), the insect body axis ratios are 2, 3, 4, and 5 respectively. Under the same body length, for insects with smaller body axis ratios, plays a leading role, the collinear polarization pattern appears to be a "cross" shape; for insects with a relatively large body axis, Plays a leading role, the longer the collinear polarization pattern is along the preset insect body length direction, based on this, the above application and It can reflect the body axis ratio of insects.

[0065] In an embodiment of the present application, multiple insect body axis ratio estimators are determined based on the association relationship, specifically including:

[0066] According to the association relationship, when the angle is 0 or When As the RCS in the body length direction. According to the association relationship, when the angle is or When As the RCS in the body width direction. According to the correlation relationship, the ratio of the RCS in the body length direction to the RCS in the body width direction is As the RCS body axis ratio indicator. 、 , RCS in body length direction, RCS in body width direction and RCS body axis ratio indicators are used as insect body axis ratio estimators, respectively.

[0067] From the above Figure 2 It can be seen that the larger the body axis ratio, the greater the difference in RCS between the body length and body width. In this case, according to the above angle setting method, this application obtains the RCS in the body length direction, the RCS in the body width direction, and the ratio of the two (RCS body axis ratio index). At this point, the server has obtained five collinear polarization pattern features related to the insect's body axis ratio, namely: 、 、 、 、 , and used them as estimators of insect body axis ratios.

[0068] S102: The server constructs a pre-trained body axis ratio inversion model based on the body axis ratio estimator of each insect and the preset XGBoost regression algorithm.

[0069] Among them, the body axis ratio inversion model establishes a mapping relationship between each insect body axis ratio estimator and the insect body axis ratio.

[0070] In the embodiment of the present application, a pre-trained axial ratio inversion model is constructed based on the axial ratio estimator of each insect and the preset extreme gradient boosting algorithm (XGBoost), which specifically includes:

[0071] Based on each insect's axis ratio estimator and several insect collinear polarization pattern samples, a number of multidimensional feature vector samples are constructed. These multidimensional feature vector samples include the multidimensional feature vectors composed of the insect axis ratio estimators corresponding to each insect's collinear polarization pattern sample, as well as the insect axis ratio labels. Each multidimensional feature vector sample is input into a preset XGBoost regression algorithm to train the algorithm until the training end condition is met, resulting in an axis ratio inversion model.

[0072] That is to say, the present application combines the advanced XGBoost regression algorithm to realize the insect body axis ratio inversion based on the above-mentioned multiple insect body axis ratio estimators. The present application performs iterative training of the XGBoost regression algorithm through a plurality of pre-set multidimensional feature vector samples. The plurality of multidimensional feature vector samples are extracted from a plurality of insect collinear polarization pattern samples to form a multidimensional feature vector of the insect body axis ratio estimator, and an expert or other equipment is used to label each multidimensional feature vector with an insect body axis ratio label. The multidimensional feature vector is as follows [ ].

[0073] Among them, the preset XGBoost regression algorithm is trained until the training end conditions are met to obtain the body axis ratio inversion model, which specifically includes:

[0074] The preset XGBoost regression algorithm is iteratively trained using the multidimensional feature vector samples in the preset training set. The multidimensional feature vector samples in the preset test set are input into the iteratively trained preset XGBoost regression algorithm to calculate the corresponding average relative error based on the output results. If the average relative error is less than a preset threshold, the preset XGBoost regression algorithm after iterative training is determined to be the body axis ratio inversion model.

[0075] When iteratively training the preset XGBoost regression algorithm in this application, the multidimensional feature vector samples may include a 75% training set and a 25% test set. For example, for insects measured at 16.2 GHz in the Ku band, 75% of the 159 insects measured in a darkroom are randomly selected to generate the training set for training the axis ratio inversion model, and the remaining 25% are used as the test set to evaluate the performance of the trained axis ratio inversion model.

[0076] This application uses the mean relative error (MRE) as the evaluation indicator of the body axis ratio estimation error:

[0077]

[0078] in, represents the number of insects involved in the inversion, Indicates the first The insect body axis ratio obtained by inversion of the insects is Indicates the first The true insect body axis ratio of each insect; The larger the value is, the greater the estimation error is. When the value is less than the preset threshold, the body axis ratio inversion model is obtained. The preset threshold is set by the user in actual use and is not specifically limited in this application.

[0079] In this embodiment of the present application, the evaluation results of the body axis ratio inversion model constructed above are shown in Table 1 (Model Evaluation Results Table). The difference in the MRE between the training and test sets can be used to determine whether the model is overfitting. Generally speaking, if the MRE of the test set is significantly greater than that of the training set, it indicates that the model is at risk of overfitting. Table 1 shows that the MRE of the model training and test sets are 9.31% and 10.57%, respectively, indicating that the model has good generalization performance. Figure 3 The figure below shows a comparison of the inverted insect body axis ratios (i.e., estimated body axis ratios) and the true body axis ratios for the 159 insects. The dots represent the insects, and the lines represent the contour lines between the estimated and true values. The body axis ratio inversion model was trained and validated based on Ku-band (16.2 GHz) data from anechoic chamber measurements of 159 insects. This is for illustrative purposes only. Model training samples can be selected in practice, such as by obtaining them online. This is not a specific limitation in this application.

[0080] Table 1 Model evaluation results

[0081]

[0082] S103: The server deploys the body axis ratio inversion model to the user terminal, so that the user terminal performs insect body axis ratio inversion based on the input insect collinear polarization pattern and the body axis ratio inversion model.

[0083] After obtaining the above-mentioned body axis ratio inversion model, the body axis ratio inversion model can be deployed on the user terminal, directly on the server, or directly locally if the execution subject is the user terminal. The user terminal can be understood as a user's mobile phone, computer, or other device, which is not specifically limited in this application.

[0084] In the embodiment of the present application, after deploying the body axis ratio inversion model to the user terminal, the method further includes:

[0085] Obtain the insect body axis ratio estimator corresponding to the input insect collinear polarization pattern and construct the corresponding multidimensional feature vector. Send the multidimensional feature vector to the user terminal so that the user terminal can perform insect body axis ratio inversion.

[0086] If the execution subject is not a user terminal but a server, the server can accept the input insect collinearity plan direction map from the user terminal or from other devices, and obtain the body axis ratio estimator of each insect according to the above steps S101-S102, and then obtain a multi-dimensional feature vector.

[0087] In another embodiment of the present application, the user terminal performs insect body axis ratio inversion based on the input insect collinear polarization pattern and body axis ratio inversion model, specifically including:

[0088] The user terminal determines multiple corresponding insect axis ratio estimators based on the input insect collinear polarization pattern and constructs a multidimensional feature vector. The user terminal inputs the multidimensional feature vector into a pre-deployed axis ratio inversion model to perform insect axis ratio inversion.

[0089] That is to say, the execution of the above steps S101-S102 can also be arranged in the user terminal, and the user terminal constructs a multi-dimensional feature vector to complete the inversion of the insect body axis ratio.

[0090] Through the above scheme, the present application deeply analyzes the preset insect collinear polarization pattern, determines the correlation between the characteristic parameters of each pattern and the insect body axis ratio, and then determines multiple insect body axis ratio estimators. These estimators can accurately characterize the quantitative relationship between the insect polarization characteristics and the insect body axis ratio, and provide an accurate data basis for the subsequent construction of the body axis ratio inversion model, thereby greatly improving the accuracy of the insect body axis ratio inversion. The body axis ratio inversion model constructed based on the preset XGBoost regression algorithm can quickly process the input insect polarization data. Compared with traditional methods, it greatly improves the efficiency of insect body axis ratio inversion, making it possible to perform body axis ratio analysis on a large number of insect samples in a short time.

[0091] The body axis ratio inversion model is deployed to the user terminal. The user only needs to input the insect's collinear polarization pattern in the terminal to perform insect body axis ratio inversion based on the model. This design gets rid of the traditional measurement method's dependence on complex equipment and professional environment, allowing insect researchers to conveniently analyze insect body axis ratios in a variety of scenarios such as the field and the laboratory, significantly enhancing the ease of application of the technology. This application effectively solves the technical problem of the difficulty in achieving efficient and accurate inversion of insect body axis ratios, providing a more accurate, efficient and convenient research method for insect research.

[0092] Figure 4 A schematic diagram of the structure of an insect body axis ratio inversion device based on RCS characteristics provided in an embodiment of the present application is shown in FIG. Figure 4 As shown, the equipment includes:

[0093] At least one processor; and a memory communicatively connected to the at least one processor. The memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to:

[0094] Based on the analysis results of the preset insect collinear polarization pattern, the correlation between the characteristic parameters of each pattern and the insect's body axis ratio is determined, so as to determine multiple insect body axis ratio estimators based on the correlation. The correlation relationship at least includes the correlation with the insect body length in the insect body axis ratio and the correlation with the insect body width in the insect body axis ratio determined based on the co-polarization radar scattering cross section RCS. The insect body axis ratio estimator is used to characterize the quantitative relationship between the insect polarization characteristics and the insect body axis ratio. According to each insect body axis ratio estimator and the preset XGBoost regression algorithm, a pre-trained body axis ratio inversion model is constructed. The body axis ratio inversion model establishes a mapping relationship between each insect body axis ratio estimator and the insect body axis ratio. The body axis ratio inversion model is deployed to the user terminal so that the user terminal performs insect body axis ratio inversion based on the input insect collinear polarization pattern and the body axis ratio inversion model.

[0095] The embodiment of the present application further provides a non-volatile computer storage medium storing computer-executable instructions, wherein the computer-executable instructions are configured as follows:

[0096] Based on the analysis results of the preset insect collinear polarization pattern, the correlation between the characteristic parameters of each pattern and the insect's body axis ratio is determined, so as to determine multiple insect body axis ratio estimators based on the correlation. The correlation relationship at least includes the correlation with the insect body length in the insect body axis ratio and the correlation with the insect body width in the insect body axis ratio determined based on the co-polarization radar scattering cross section RCS. The insect body axis ratio estimator is used to characterize the quantitative relationship between the insect polarization characteristics and the insect body axis ratio. According to each insect body axis ratio estimator and the preset XGBoost regression algorithm, a pre-trained body axis ratio inversion model is constructed. The body axis ratio inversion model establishes a mapping relationship between each insect body axis ratio estimator and the insect body axis ratio. The body axis ratio inversion model is deployed to the user terminal so that the user terminal performs insect body axis ratio inversion based on the input insect collinear polarization pattern and the body axis ratio inversion model.

[0097] The various embodiments in this application are described in a progressive manner. Similar parts between the various embodiments can be referred to in conjunction with each other. Each embodiment focuses on the differences between the other embodiments. In particular, the device and medium embodiments are generally similar to the method embodiments, so the description is relatively simple. For relevant parts, refer to the description of the method embodiments.

[0098] The devices and media provided in the embodiments of the present application correspond one-to-one to the methods. Therefore, the devices and media also have similar beneficial technical effects to their corresponding methods. Since the beneficial technical effects of the methods have been described in detail above, the beneficial technical effects of the devices and media will not be repeated here.

[0099] It should also be noted that the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, commodity, or apparatus that includes a series of elements includes not only those elements but also other elements not explicitly listed, or includes elements inherent to such process, method, commodity, or apparatus. In the absence of further limitations, an element defined by the phrase "comprises a ..." does not exclude the presence of other identical elements in the process, method, commodity, or apparatus that includes the element.

[0100] The foregoing is merely an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application should all be included within the scope of the claims of the present application.

Claims

1. A method for inverting insect body axis ratio based on RCS features, characterized by: The method comprises: Based on the analysis results of preset insect collinear polarization patterns, a correlation relationship between characteristic parameters of each pattern and the insect's body axis ratio is determined, and multiple insect body axis ratio estimators are determined based on the correlation relationship; wherein the correlation relationship includes at least a correlation with the insect body length in the insect body axis ratio and a correlation with the insect body width in the insect body axis ratio determined based on the co-polarization radar scattering cross section (RCS); the insect body axis ratio estimator is used to characterize the quantitative relationship between the insect polarization characteristics and the insect body axis ratio; According to each of the insect body axis ratio estimators and the preset XGBoost regression algorithm, a pre-trained body axis ratio inversion model is constructed; wherein the body axis ratio inversion model establishes a mapping relationship between each of the insect body axis ratio estimators and the insect body axis ratio; Deploying the body axis ratio inversion model to a user terminal, so that the user terminal performs insect body axis ratio inversion based on an input insect collinear polarization pattern and the body axis ratio inversion model; Wherein, determining a plurality of insect body axis ratio estimators based on the association relationship specifically includes: According to the relationship, when the angle is 0 or When As the body length direction RCS; the association relationship includes Related to the elongation of the collinear polarization pattern along the body length, The collinear polarization pattern is correlated along both the insect's body length and width; and is a pattern characteristic parameter related to the shape of the insect collinear polarization pattern; the angle is used to determine the preset insect collinear polarization pattern. Item and When the items respectively reach their maximum values, the angle between the polarization direction and the insect body axis direction; According to the association relationship, when the angle is or When As the RCS in the body width direction; According to the association relationship, the ratio of the body length direction RCS to the body width direction RCS As an indicator of RCS body axis ratio; The 、 , the RCS in the body length direction, the RCS in the body width direction and the RCS body axis ratio index, respectively serving as the insect body axis ratio estimator; Wherein, before determining the correlation between the characteristic parameters of each pattern and the insect body axis ratio based on the analysis results of the preset insect collinear polarization pattern, the method further includes: Obtaining co-polarization radar cross sections (RCS) corresponding to different polarization directions of a pre-observed insect; wherein the co-polarization RCS is obtained by observing the pre-observed insect using a polarization insect radar in a linear polarization mode; The preset insect collinear polarization pattern is determined according to the change of the co-polarization RCS in the polarization direction; wherein, under the assumption that the insect is symmetric, the preset insect collinear polarization pattern is expressed as: in, represents the insect collinear polarization pattern, is the characteristic parameter of the pattern that is independent of the polarization direction; Indicates the polarization direction, and its value range is ; Indicates the direction of the insect's body axis; Among them, based on the analysis results of the preset insect collinear polarization pattern, the correlation between the characteristic parameters of each pattern and the insect body axis ratio is determined, specifically including: According to the angle side directions corresponding to the angles and the preset insect body length direction, respectively determining the body length direction and body width direction corresponding to the polarization directions; According to the Item and The corresponding relationship between the item and the body length direction and the body width direction is established. and stated The correlation relationship between the insect body axis ratio and the insect body axis ratio; According to the insect body axis ratio estimator and the preset XGBoost regression algorithm, a pre-trained body axis ratio inversion model is constructed, specifically including: According to each of the insect body axis ratio estimators and a number of insect collinear polarization pattern samples, a number of multidimensional feature vector samples are constructed; wherein, the multidimensional feature vector samples include a multidimensional feature vector composed of each of the insect body axis ratio estimators corresponding to each of the insect collinear polarization pattern samples and an insect body axis ratio label; the multidimensional feature vector is [ ]; Inputting each of the multidimensional feature vector samples into the preset XGBoost regression algorithm to train the preset XGBoost regression algorithm until a training end condition is met, thereby obtaining the body axis ratio inversion model; The user terminal performs insect body axis ratio inversion based on the input insect collinear polarization pattern and the body axis ratio inversion model, specifically including: The user terminal determines a plurality of corresponding insect body axis ratio estimators according to the input insect collinear polarization pattern, and constructs the multidimensional feature vector; The user terminal inputs the multi-dimensional feature vector into the pre-deployed body axis ratio inversion model to perform insect body axis ratio inversion.

2. The insect body axis ratio inversion method based on RCS characteristics according to claim 1 is characterized in that: The preset XGBoost regression algorithm is trained until the training end condition is met to obtain the body axis ratio inversion model, specifically including: Iteratively training the preset XGBoost regression algorithm through each of the multidimensional feature vector samples in the preset training set; Inputting each of the multidimensional feature vector samples in the preset test set into the preset XGBoost regression algorithm after iterative training to calculate the corresponding average relative error according to the output result; When the average relative error is less than a preset threshold, the preset XGBoost regression algorithm after iterative training is determined as the body axis ratio inversion model.

3. The insect body axis ratio inversion method based on RCS characteristics according to claim 1 is characterized in that: After deploying the body axis ratio inversion model to the user terminal, the method further includes: Obtaining each of the insect body axis ratio estimators corresponding to the input insect collinear polarization pattern, and constructing the corresponding multidimensional feature vector; The multidimensional feature vector is sent to the user terminal, so that the user terminal performs insect body axis ratio inversion.

4. An insect body axis ratio inversion device based on RCS characteristics, characterized in that: The device comprises: at least one processor; and, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the insect body axis ratio inversion method based on RCS features as described in any one of claims 1 to 3.

5. A non-volatile computer storage medium storing computer-executable instructions, characterized in that: The computer executable instructions can execute an insect body axis ratio inversion method based on RCS features as described in any one of claims 1 to 3.

Images