Forest resource prediction method based on remote sensing image analysis

Abstract

The invention relates to the technical field of image processing, and discloses a forest resource prediction method based on remote sensing image analysis, which comprises the following steps: acquiring a multi-source remote sensing image, performing radiation and geometric correction, and acquiring a registered multi-source remote sensing image through image registration; performing wavelet decomposition on the registered multi-source remote sensing image, performing weighted fusion on low-frequency components, performing fusion on high-frequency components based on regional energy, and performing wavelet inverse transformation to generate a fused remote sensing image; extracting spectral features of the fused remote sensing image, and generating a spectral feature matrix; taking the spectral feature matrix as feature variables, taking actual observation values of forest resources as prediction target variables, calculating mutual information values between the feature variables and the prediction target variables to screen optimal spectral features, and finally realizing forest resource prediction by calculating a Gini index and an information gain and constructing an improved random forest model. According to the invention, the precision of the forest resource prediction result is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of image processing, and particularly relates to a forest resource prediction method based on remote sensing image analysis. BACKGROUND

[0002] Forest resources are important ecological resources on the earth, and play an important role in maintaining ecological balance, providing timber and other forest products, and regulating climate. Accurate prediction of the change of forest resources is of great significance for the scientific management, sustainable utilization and ecological environment protection of forest resources.

[0003] Traditional forest resource prediction methods mainly rely on ground surveys. Although this method can obtain relatively accurate local data, it has the disadvantages of time-consuming, labor-intensive, high cost, limited coverage, etc., and it is difficult to meet the needs of large-scale, rapid and dynamic forest resource prediction.

[0004] With the development of remote sensing technology, remote sensing images have been widely used in forest resource monitoring due to their wide coverage, short acquisition period and rich information. However, most of the current forest resource prediction methods based on remote sensing images are based on a single data source, which has the problems of incomplete data information and low prediction accuracy. At the same time, the existing methods are difficult to fully mine the effective information in the multi-source remote sensing image data, resulting in unsatisfactory prediction results. SUMMARY

[0005] In view of the above problems in the prior art, the present application provides a forest resource prediction method based on remote sensing image analysis, which solves the problem of low forest resource prediction accuracy caused by the single data source, incomplete information and poor method processing effect of the existing forest resource prediction method.

[0006] In order to achieve the above-mentioned application purposes, the technical scheme adopted by the present application is as follows: A forest resource prediction method based on remote sensing image analysis, comprising the following steps: S1, obtaining optical remote sensing images and radar remote sensing images of a target forest area, performing data screening, and obtaining screened multi-source remote sensing images; S2, after the screened multi-source remote sensing images are radiometrically and geometrically corrected, image registration is performed to generate registered multi-source remote sensing images; S3, based on the registered multi-source remote sensing images, wavelet decomposition is performed on each remote sensing image to extract low-frequency components and high-frequency components, the low-frequency components are weighted and fused, and the high-frequency components are fused based on regional energy, and after the fused low-frequency components and high-frequency components are generated, inverse wavelet transform is performed to generate fused remote sensing images; S4, based on the fused remote sensing image, extracting the spectral features of forest resources to generate a spectral feature matrix; S5, taking the spectral feature matrix as a characteristic variable and the actual observation value of the forest resources as a prediction target variable, calculating the mutual information value between each characteristic variable and the prediction target variable, screening the characteristic variables with mutual information values greater than the mutual information threshold, and generating an optimized spectral feature; S6, taking the optimized spectral feature as input, constructing an improved multiple decision tree by calculating the Gini index and information gain, and generating an improved random forest model; S7, reacquiring the spectral features of the target forest area and inputting them into the improved random forest model to generate a forest resource prediction result.

[0007] The present application has the following beneficial effects: 1. The forest resource prediction method based on remote sensing image analysis provided by the present application comprehensively utilizes the advantages of different data sources by collecting multi-source remote sensing image data, provides more comprehensive and rich information, and thus improves the accuracy of forest resource prediction. 2. After data processing on the multi-source remote sensing image, image registration and data fusion are performed after radiation correction and geometric correction, which not only effectively eliminates errors and redundant features in the data, but also improves the quality and consistency of the data. 3. The improved random forest model is constructed by optimizing feature selection and decision tree construction, which improves the performance and prediction accuracy of the improved random forest model, so that the improved random forest model can better adapt to the complex scene of forest resource prediction. BRIEF DESCRIPTION OF DRAWINGS

[0008] Figure 1 The flowchart of the forest resource prediction method based on remote sensing image analysis provided by the present application is shown. DETAILED DESCRIPTION

[0009] The specific embodiments of the present application are described below to facilitate understanding of the present application by those skilled in the art, but it should be clear that the present application is not limited to the scope of the specific embodiments. It is obvious to those skilled in the art that all the applications utilizing the concept of the present application are within the scope of the present application as defined and limited by the appended claims.

[0010] As shown in Figure 1 A forest resource prediction method based on remote sensing image analysis includes the following steps S1-S7: S1, acquiring optical remote sensing images and radar remote sensing images of a target forest area, performing data screening, and obtaining screened multi-source remote sensing images.

[0011] Specifically, step S1 specifically comprises S11-S13: S11, determine the latitude and longitude range of the target forest area and the time span of data acquisition.

[0012] S12, obtain optical remote sensing images and radar remote sensing images of the target forest area.

[0013] S13, data screening is performed on the optical remote sensing images and the radar remote sensing images, and the remote sensing images with large-area point cloud coverage and data damage are removed, to generate screened multi-source remote sensing images.

[0014] In this embodiment, this step is the process of obtaining multi-source remote sensing images, specifically: determining the latitude and longitude range of the target forest area, and determining the time span of data acquisition, such as nearly 5 years, once every quarter; obtaining optical remote sensing images through a remote sensing satellite data platform, such as a Landsat series or a Sentinel series satellite data platform; obtaining radar remote sensing images through a radar satellite data platform; screening the obtained data to remove images with problems such as large-area cloud coverage and data damage, to ensure that the obtained multi-source remote sensing image data completely covers the target forest area in space and forms a continuous sequence in time, so as to build a high-quality, multi-dimensional data basis for subsequent forest resource prediction.

[0015] S2, after radiation correction and geometric correction of the screened multi-source remote sensing images, image registration is performed to generate registered multi-source remote sensing images.

[0016] Specifically, step S2 specifically comprises S21-S22: S21, radiation correction and geometric correction are performed on the screened multi-source remote sensing images to generate corrected multi-source remote sensing images.

[0017] In this embodiment, the purpose of performing radiation correction and geometric correction on the screened multi-source remote sensing images is to eliminate interference errors of the multi-source remote sensing images, and to provide accurate and consistent data basis for subsequent forest resource related analysis (such as registration, feature extraction, and prediction).

[0018] The process of radiation correction is as follows: for optical remote sensing images, collect atmospheric parameters during image acquisition, including atmospheric aerosol concentration, water vapor content, etc., select a suitable atmospheric model, input the digital quantization value (DN value) of the optical remote sensing image into the atmospheric transmission model, and calculate the reflectivity of the ground surface to complete the radiation correction; for radar remote sensing images, according to the parameters of the radar sensor, including radar wavelength, incident angle, etc., the radar echo intensity is converted into backscattering coefficient by using the radiation calibration formula to realize the radiation correction; therefore, the radiation correction can remove the interference of non-target factors such as atmospheric scattering and sensor error on the image gray value, and restore the real radiation information of the ground object (such as forest vegetation), so as to ensure that the gray features of the same ground object in different images are consistent, and provide a reliable basis for subsequent extraction of key features (support resource prediction) such as forest spectrum or structure.

[0019] The process of geometric correction is as follows: on the remote sensing image after radiation correction, at least 20 ground control points are uniformly selected, and the actual geographic coordinates, i.e. latitude and longitude coordinates, of these control points are obtained; a quadratic polynomial fitting method is used to establish the mathematical relationship between the pixel coordinates of the remote sensing image and the actual geographic coordinates, and the polynomial coefficients are solved by iterative calculation; the remote sensing image is resampled (interpolated) by using the solved polynomial coefficients to obtain the remote sensing image after geometric correction; therefore, the geometric correction can eliminate the spatial position deviation of the image caused by satellite attitude, terrain undulation, etc., ensure the accurate alignment of remote sensing images of different sources and different time phases in spatial coordinates, and avoid the influence of position misalignment on the accurate identification of forest area range and the feature comparison across images.

[0020] S22, the corrected multi-source remote sensing images are image registered to generate registered multi-source remote sensing images, specifically: S221, from the corrected multi-source remote sensing images, a remote sensing image with high definition and resolution is randomly selected as a reference image, and the remaining images are taken as images to be registered.

[0021] In this embodiment, the purpose of selecting a remote sensing image with high definition and resolution is to make the selected reference image have high-quality, clear ground feature and accurate geometric positioning, so as to improve the registration accuracy of the image.

[0022] S222, after the reference image and the image to be registered are grayed, the scale-invariant feature transformation method is used to extract the feature points of the reference image and the image to be registered.

[0023] S223, based on the feature points of the reference image and the image to be registered, the fast nearest neighbor search method is used to perform approximate nearest neighbor matching between the feature points of the reference image and the image to be registered, to obtain the feature points in the image to be registered that are most similar to the feature points of the reference image, and to generate a pair of preliminary matched feature points.

[0024] In this embodiment, for the feature point set of the reference image, a fast nearest neighbor search index is constructed, for each feature point of the image to be registered, the nearest neighbor and the second nearest neighbor feature points in the index of the reference image are searched, and then the Lowe ratio test is used to screen the matching pairs, that is, if the ratio of the distance between the nearest feature point and the second nearest feature point is less than a set value (0.75), the matching pair is retained, and finally the preliminary matched feature point pairs are obtained. and the second nearest feature point .

[0025] S224, based on the preliminary matched feature point pairs, a random sample consensus method is used for detection, an affine transformation model is constructed, and the false matched feature point pairs are removed to generate the optimized feature point pairs.

[0026] In this embodiment, according to the preliminary matched feature point pairs, 3 matching pairs are randomly selected from the preliminary matching pairs, and an affine transformation matrix is calculated, that is:

[0027] wherein, is the homogeneous coordinates of the feature points of the reference image, , are the horizontal coordinates and the vertical coordinates of the feature points in the reference image respectively, and 1 is the normalized form of the homogeneous coordinates, which is used for unified processing of translation, rotation, scaling and other linear transformations; is the homogeneous coordinates of the feature points of the image to be registered, , are the horizontal coordinates and the vertical coordinates of the feature points in the image to be registered respectively, and 1 is also the normalized element of the homogeneous coordinates; is the affine transformation matrix, which is a 3x3 matrix, and the form is which contains the translation, rotation, scaling, shearing and other transformation information of the image to be registered to the reference image, and through the matrix, the points in the image to be registered can be mapped to the corresponding positions in the reference image, , , , , , all represent the elements of the affine transformation matrix, which together determine the specific affine transformation mode, which is calculated through the randomly selected matching pairs; Then, all the matching pairs are transformed through the affine transformation matrix calculated, and the Euclidean distance between the predicted coordinates after transformation and the actual coordinates is calculated that is:

[0028] wherein, , These are the predicted x and y coordinates of the feature points in the image to be registered in the reference image after transformation by the affine transformation matrix; if the Euclidean distance... If the value is less than a set threshold (3 pixels in this invention), the matching pair is an interior point (a correct matching point pair); otherwise, it is an exterior point, i.e., an incorrect matching point pair. In addition, iterative optimization is required, i.e., repeated random sampling and continuing the above interior point judgment process. When the maximum number of iterations is reached (1000 times in this invention), the affine transformation matrix with the most interior points and the corresponding interior point set are retained. This interior point set is the optimized feature point pair.

[0029] S225. Based on the optimized feature point pairs, calculate the affine transformation matrix from the image to be registered to the reference image.

[0030] S226. Use the affine transformation matrix to perform spatial transformation on each pixel of the image to be registered, and obtain the new coordinates of the image to be registered in the coordinate system of the reference image, and finally generate the registered multi-source remote sensing image. In this embodiment, after spatial transformation, the transformed pixel coordinates may not be integers. Therefore, in practical applications, it is necessary to resample the new coordinates generated after spatial transformation. For example, the bilinear interpolation method can be selected to interpolate the new coordinates generated after transformation, and finally the registered multi-source remote sensing image is obtained.

[0031] S3. Based on the registered multi-source remote sensing images, wavelet decomposition is performed on each remote sensing image to extract low-frequency and high-frequency components. The low-frequency components are weighted and fused, while the high-frequency components are fused based on regional energy. After generating the fused low-frequency and high-frequency components, inverse wavelet transform is performed to generate the fused remote sensing image.

[0032] In this embodiment, multi-source remote sensing images (optical and radar images) have different information advantages. Optical images have rich spectral information, while radar images can penetrate clouds and fog and reflect terrain and landform structures well. Therefore, by wavelet decomposition, the low-frequency (reflecting the overall image overview, main energy and contour information) and high-frequency (reflecting image details, edges, textures, etc.) components of different images are fused separately. This can integrate the advantages of multi-source images, so that the fused image retains the overall clear outline and has rich detailed features, and enhances the ability to express information about forest resources (such as vegetation type, tree distribution, growth status, etc.). At the same time, the fused remote sensing image integrates the effective information of multi-source images, and has higher quality and more comprehensive information. This provides a more accurate and richer data foundation for subsequent forest resource prediction based on remote sensing images (such as vegetation cover monitoring), which helps to improve the accuracy and reliability of prediction.

[0033] Specifically, step S3 includes S31-S34: S31, set the wavelet function and the decomposition layer number, and perform wavelet decomposition on each registered remote sensing image to obtain a low-frequency component and a high-frequency component of each remote sensing image.

[0034] In this embodiment, the wavelet function is a db4 wavelet function, and the decomposition layer number is 3 layers.

[0035] S32, based on the low-frequency component, calculate the definition index of the low-frequency component of each remote sensing image, determine the weight of the low-frequency component, generate a fused low-frequency component through weighted fusion, and specifically: S321, calculate the average gradient of the low-frequency component of each remote sensing image, and take it as the definition index, that is:

[0036] wherein, indicates the average gradient of the low-frequency component of the i-th remote sensing image, which is used to measure the definition of the image region corresponding to the low-frequency component, and the greater the average gradient, the clearer the image, , , respectively indicate the number of pixels of the low-frequency component along the horizontal and vertical directions, indicates the gray value of the low-frequency component of the i-th remote sensing image at the pixel point, indicates the gray value of the low-frequency component of the i-th remote sensing image at the pixel point, indicates the gray value of the low-frequency component of the i-th remote sensing image at the pixel point, indicates the gray value of the low-frequency component of the i-th remote sensing image at the pixel point, indicates the gray value of the low-frequency component of the i-th remote sensing image at the pixel point. S322, sort the average gradient of the low-frequency component of each remote sensing image from large to small, set the weight of the remote sensing image with the larger average gradient to be larger, and the sum of the weights of each remote sensing image is 1, to obtain the weight corresponding to the average gradient of the low-frequency component of each remote sensing image. S323, perform weighted average on the average gradient of the low-frequency component of each remote sensing image and the weight corresponding thereto to obtain a fused low-frequency component, that is:

[0037]

[0038] wherein, indicates the fused low-frequency component, indicates the number of remote sensing images, indicates the weight of the i-th remote sensing image.

[0039]

[0040] ​​​​​In this embodiment, the step is to generate the fused low-frequency component by performing weighted fusion based on average gradient on the low-frequency components of the multi-source remote sensing images, so that not only the main information of the clear image can be retained, but also the fusion quality is optimized, that is, the average gradient is used to reflect the definition of the image, and the low-frequency component of the remote sensing image with larger average gradient (that is, the image is clearer) is given larger weight, in the fusion process, the key information such as the main contour and overall structure contained in the clear image is retained more in the fused low-frequency component, so that the main features of the fused image are clear and identifiable, at the same time, the low-frequency components of the multiple remote sensing images are integrated by the weighted average method, the information of the multi-source images is integrated, the information loss or blur problem of a single image is avoided, the fused low-frequency component is more accurate and complete in overall appearance, which lays a foundation for generating high-quality fused remote sensing image, and then improves the accuracy of subsequent forest resource prediction based on the fused image.

[0041] S33, performing region energy-based fusion on the high-frequency components to generate a fused high-frequency component, specifically: S331, based on the high-frequency components, calculating the energy value of an n*n region around each pixel point of each remote sensing image, that is:

[0042] wherein, Ei (x, y) represents the energy value of the high-frequency component of the i-th remote sensing image in the n*n region around the pixel point (x, y), Ei (x, y) represents the energy value of the high-frequency component of the i-th remote sensing image in the n*n region around the pixel point (x, y), Ei (x, y) represents the high-frequency coefficient of the i-th remote sensing image at the pixel point (x, y), which reflects the high-frequency information of the image at the pixel point, such as edge, texture and other detail features, n represents the size of the region around the pixel point, represents a down rounding operation. In this embodiment, the value of n of the present application is 3, but in actual application, the value of n can be determined according to the image resolution and the detail scale, for example, when the resolution of the remote sensing image is 10m, n can be 3, 5 or other odd numbers, so as to ensure the field center symmetry. S332, for each pixel point, comparing the energy values of all remote sensing images at the pixel point, selecting the remote sensing image with the maximum energy value, taking the high-frequency coefficient of the remote sensing image as the optimal high-frequency coefficient of the pixel point, and finally splicing the optimal high-frequency coefficients of all pixel points to generate a fused high-frequency component, that is:

[0043]

[0044]

[0045] ​​​

[0046] wherein, denotes the coefficient of the fused high-frequency component at the pixel point , denotes the high-frequency coefficient of the remote sensing image at the pixel point , denotes the remote sensing image at the pixel point , denotes the independent variable obtained when the function takes the maximum value, that is, the remote sensing image .

[0047] In this embodiment, this step generates the fused high-frequency component by fusing the high-frequency components based on the regional energy. Not only can the details information be accurately retained and the image detail expression ability be enhanced, but also the reliability of subsequent analysis is improved, that is, the high-frequency component carries the details, edges, textures and other information of the image. When fused based on the regional energy, the high-frequency coefficient of the remote sensing image with the maximum regional energy around each pixel point is selected, so that the most detailed part of the multi-source remote sensing image can be accurately retained, and the fused high-frequency component contains rich and clear edge, texture and other detail features, such as the outline of trees in the forest and the subtle distribution difference of vegetation. At the same time, by integrating the part with the maximum regional energy in the high-frequency component of the multi-source remote sensing image, the fused high-frequency component is enhanced in the richness and clarity of details, so that the subsequently generated fused remote sensing image is more visually layered and the details are more prominent, which helps to more accurately identify and analyze the subtle features of forest resources, such as the growth state of trees and the texture change of vegetation caused by pests and diseases. In addition, the high-quality high-frequency component fusion provides a better detailed data basis for subsequent forest resource prediction and other work based on the fused remote sensing image, so that related analysis (such as forest stock volume estimation and vegetation coverage monitoring) can more accurately capture key details, thereby improving the reliability and accuracy of analysis and prediction results.

[0048] S34, inverse wavelet transform is performed on the fused low-frequency component and the high-frequency component to generate a fused remote sensing image.

[0049] S4, based on the fused remote sensing image, a spectral feature of the forest resource is extracted to generate a spectral feature matrix.

[0050] Specifically, step S4 specifically includes S41-S42: S41, based on the fused remote sensing image, the reflectance values of the near-infrared band and the red band are extracted, and the normalized vegetation index value of each pixel point is calculated as a spectral feature, that is:

[0051] wherein, This represents the normalized vegetation index value for each pixel. Represents the reflectivity in the near-infrared band. This indicates the reflectivity in the red light band.

[0052] In this embodiment, the reflectance calculation of the infrared and red light bands can effectively reflect information such as vegetation coverage and growth status, and can be used as the spectral characteristics of forest resources for subsequent forest resource prediction.

[0053] S42. The normalized vegetation index values ​​of each pixel are concatenated to generate a spectral feature matrix.

[0054] In this embodiment, NDVI is used to effectively reflect the characteristics of vegetation cover and growth vitality, providing basic data for subsequent calculation of the mutual information value between feature variables (NDVI values) and prediction target variables (such as forest cover), so as to screen out spectral features that are more valuable for forest cover prediction, and lay a key feature data foundation for building a random forest model and accurately predicting forest resources.

[0055] S5. Using the spectral feature matrix as the feature variable and the actual observed value of forest resources as the prediction target variable, calculate the mutual information value between each feature variable and the prediction target variable, filter the feature variables with mutual information values ​​greater than the mutual information threshold, and generate the optimized spectral features.

[0056] Specifically, actual observed values ​​of forest resources include forest coverage.

[0057] Specifically, the formula for calculating the mutual information value between each feature variable and the target variable in step S5 is as follows:

[0058] in, Representing characteristic variables With the target variable for prediction The mutual information value between a feature variable and the target variable quantifies how much information they share. A larger value indicates a stronger correlation between the two. Therefore, using mutual information values ​​to filter spectral features can retain the most effective features for forest resource prediction, such as forest cover prediction. Representing characteristic variables, Indicates the target variable to be predicted. Representing characteristic variables With the target variable for prediction The joint probability distribution, Represents the logarithmic function. Representing characteristic variables Marginal probability distribution, Indicates the target variable for prediction The marginal probability distribution.

[0059] In this embodiment, the mutual information threshold is determined by an unsupervised method, that is, the median of the mutual information is taken as the threshold, thereby filtering out low-relevance features.

[0060] In summary, this step, by calculating the mutual information value between spectral feature variables and actual forest resource observations and then filtering features, not only accurately selects effective features but also improves the prediction accuracy and efficiency of subsequent models, while enhancing feature interpretability. Specifically, mutual information quantifies the degree of information sharing between spectral features and actual forest resource observations, filtering out feature variables with mutual information values ​​greater than a threshold. These features are more strongly correlated with actual forest resource observations, accurately retaining the most valuable spectral information for predicting actual forest resource observations and avoiding irrelevant or weakly correlated features from interfering with subsequent model construction. Furthermore, removing redundant and irrelevant spectral features reduces the number of input features for subsequent random forest models, lowering the computational complexity of model training and improving training efficiency. Building models based on more effective features helps improve the accuracy and reliability of forest resource predictions. In addition, the features selected through mutual information have a clearer correlation with actual forest resource observations, enhancing the interpretability of prediction results in subsequent random forest models built based on these features and facilitating the analysis of the influence mechanism of spectral features on actual forest resource observations.

[0061] S6. Using the optimized spectral features as input, and by calculating the Gini index and information gain, construct multiple improved decision trees to generate an improved random forest model.

[0062] In this embodiment, this step improves the performance of individual decision trees and constructs an improved random forest model, thereby enhancing the predictive power of the random forest model. Specifically, during the node splitting process of the decision tree, the Gini index (measuring node purity) and information gain (measuring the contribution of features to classification) are comprehensively considered, and a weighted average is calculated using the analytic hierarchy process (AHP) to select the optimal splitting feature. This makes the feature selection of individual decision trees more accurate, better capturing the complex relationship between spectral features and forest resources, and improving the classification or regression performance of individual decision trees. Simultaneously, the random forest model, composed of multiple improved decision trees, retains the ensemble learning advantages of random forests (such as reducing overfitting risk and improving model stability). Furthermore, due to the improved performance of each decision tree, the entire random forest model has higher accuracy and generalization ability when predicting forest resources (such as forest cover), enabling it to more reliably complete forest resource prediction tasks. The process of generating the improved random forest model is as follows: Specifically, step S6 includes S61-S63: S61. Randomly sample the optimized spectral features to construct multiple decision trees.

[0063] S62. Perform feature evaluation on each decision tree and construct multiple improved decision trees, specifically as follows: S621. For each node of each decision tree, calculate the weighted Gini index and information gain for each candidate feature, i.e.:

[0064]

[0065]

[0066]

[0067]

[0068]

[0069] in, Indicates the current node The Gini index, This represents the total number of categories of samples in the current node. Indicates the first The proportion of class samples in the current node. Indicates the current node The sample set, Representing candidate features, Representing candidate features The number of possible values, Representing candidate features Values a subset of samples Represents a subset of samples The Gini index, Represents a subset of samples The Middle The proportion of class samples, Representing candidate features The weighted Gini index, Represents the sample set of the current node The information gain of a candidate feature is considered; the larger the value, the more helpful the feature is for classification. Represents the sample set Information entropy Represents a subset of samples Information entropy This represents the logarithmic function with base 2.

[0070] In this embodiment, ,and ,because Indicates the first The proportion of each class of samples in the current node; in the sample set of a node, the proportion of all classes (total) The sum of the proportions of samples from each class must equal all samples from that node (because these samples must belong to and only belong to this class). (a certain category within a category), so from the perspective of the definition of proportion and the completeness of the set, the sum of the proportions of all categories is 1. Furthermore, the current node... The formula for calculating the Gini index is as follows: Therefore, the sample subset The formula for calculating the Gini index is: Then, the proportion of the sample size in the subset is used as the weight. Candidate features can be obtained by weighted summation of the Gini indices of all child nodes. The Gini index, also known as the weighted Gini index, is: This design allows the Gini index of child nodes with a large sample size to be more prominent in candidate features. It accounts for a larger proportion of the overall Gini index and can more accurately measure the reduction in impurity of the entire dataset after splitting the parent node with candidate features, thus providing a more reasonable basis for the subsequent decision tree to select the optimal splitting feature (selecting the optimal candidate feature as the splitting feature).

[0071] S622. Using the analytic hierarchy process (AHP), weights are assigned to the weighted Gini index and information gain, and the comprehensive value of each candidate feature is calculated, i.e.:

[0072] in, This represents the combined value of each candidate feature. , These represent the weighted Gini index and information gain for a candidate feature, respectively. , These represent the weights of the weighted Gini index and the information gain, respectively.

[0073] In this embodiment, the weights of the weighted Gini index and information gain are determined using the analytic hierarchy process (AHP), specifically by constructing a decision matrix, calculating the largest eigenvalue and eigenvector, and performing a one-time test. Furthermore, in practical applications, cross-validation can also be used to select the weights of the weighted Gini index and information gain, thereby optimizing the performance of a single decision tree.

[0074] S623, select the candidate feature with the optimal comprehensive value as the split feature, realize the improvement of the single decision tree, and repeatedly execute S621-S623 to realize the improvement of multiple decision trees, and finally generate improved multiple decision trees.

[0075] In the embodiment, in the construction process of the decision tree, a split criterion based on the combination of Gini index and information gain is used to improve the classification performance of the decision tree. For each node, the weighted Gini index and information gain of each feature are calculated, then the two are combined according to a certain weight, the feature with the optimal comprehensive value is selected as the split feature, and finally the improvement of the single decision tree is realized.

[0076] S63, the improved multiple decision trees are combined into a random forest model to generate an improved random forest model.

[0077] In the embodiment, the optimized spectral feature set is divided into a training set and a test set according to a ratio of 7:3. The training set is used to train the improved random forest model, the model parameters, i.e. the number of decision trees and the maximum depth, are adjusted to make the model achieve good performance on the training set; then the test set is input into the trained model to obtain the prediction result of the forest resources; therefore, after obtaining the trained improved random forest model, the spectral features of any target forest area are input into the model, and the prediction result of the forest resources can be generated.

[0078] S7, the spectral features of the target forest area are reacquired and input into the improved random forest model to generate the prediction result of the forest resources.

[0079] The forest resource prediction method based on remote sensing image analysis provided by the application can filter out invalid data by screening multiple source remote sensing images, ensure the reliability of data input, process the screened multiple source remote sensing images, eliminate sensor, atmospheric and geometric deformation errors through radiation and geometric correction, unify the spatial coordinates of the multiple source remote sensing images through image registration to solve the data deviation problem, then perform wavelet decomposition on the registered multiple source remote sensing images, weight and fuse the low-frequency components to retain the overall information of the ground objects, fuse the high-frequency components according to the regional energy to strengthen the forest detail features, and generate the fused remote sensing image with more information through inverse transformation to provide high-quality data support for feature extraction; then the spectral features related to forest prediction resources are extracted from the fused remote sensing image to construct a matrix, the features with strong correlation with the actual observation values of the forest are selected according to the mutual information value, and thus an improved random forest model is constructed, i.e. the decision tree construction rule is optimized to improve the capture ability of the improved random forest model to the changes of the forest resources, and finally the accuracy of the forest resource prediction is improved; therefore, the error is gradually reduced and the effectiveness is strengthened from the data, feature and model levels, and finally the forest resource prediction accuracy is significantly improved.

[0080] The principles and implementation manners of the present application are described by using specific examples in the present application. The above examples are only used to help understand the method of the present application and its core idea. Meanwhile, for those skilled in the art, according to the idea of the present application, the specific implementation manners and application scopes will be changed. In summary, the content of the present description should not be understood as a limitation on the present application.

[0081] Those skilled in the art will understand that the examples described herein are for the purpose of understanding the principles of the present application and should be understood as not limiting the scope of protection of the present application. Those skilled in the art can make various other specific modifications and combinations according to the technical inspiration disclosed in the present application without departing from the essence of the present application, and these modifications and combinations are still within the scope of protection of the present application.

Claims

1. A forest resource prediction method based on remote sensing image analysis, characterized by, The method comprises the following steps: S1, acquiring optical remote sensing images and radar remote sensing images of a target forest area, performing data screening, and acquiring screened multi-source remote sensing images; S2, after radiation correction and geometric correction of the screened multi-source remote sensing images, performing image registration to generate registered multi-source remote sensing images; S3, based on the registered multi-source remote sensing images, performing wavelet decomposition on each remote sensing image, extracting low-frequency components and high-frequency components, performing weighted fusion on the low-frequency components, and simultaneously performing zone energy-based fusion on the high-frequency components, generating fused low-frequency components and high-frequency components, and then performing wavelet inverse transformation to generate fused remote sensing images; S4, based on the fused remote sensing images, extracting spectral features of forest resources to generate a spectral feature matrix; S5, taking the spectral feature matrix as a characteristic variable and the actual observation value of the forest resources as a prediction target variable, calculating mutual information values between each characteristic variable and the prediction target variable, screening characteristic variables with mutual information values greater than a mutual information threshold, and generating optimized spectral features; S6, taking the optimized spectral features as input, constructing multiple improved decision trees by calculating Gini indexes and information gains, and generating an improved random forest model; S7, reacquiring spectral features of the target forest area, inputting the spectral features into the improved random forest model, and generating a forest resource prediction result.

2. The forest resource prediction method based on remote sensing image analysis according to claim 1, characterized in that, Step S1 specifically comprises: S11, determining the latitude and longitude range of the target forest area and the time span of data acquisition; S12, acquiring optical remote sensing images and radar remote sensing images of the target forest area; S13, performing data screening on the optical remote sensing images and the radar remote sensing images, eliminating remote sensing images with large-area point cloud coverage and data damage, and generating screened multi-source remote sensing images.

3. The forest resource prediction method based on remote sensing image analysis according to claim 1, characterized in that, Step S2 specifically comprises: S21, performing radiation correction and geometric correction on the screened multi-source remote sensing images to generate corrected multi-source remote sensing images; S22, performing image registration on the corrected multi-source remote sensing images to generate registered multi-source remote sensing images, specifically: S221, from the corrected multi-source remote sensing images, randomly selecting a remote sensing image with high clarity and resolution as a reference image, and taking the remaining images as images to be registered; S222, after gray-scale processing of the reference image and the images to be registered, extracting feature points of the reference image and the images to be registered by using a scale-invariant feature transformation method; S223, based on the feature points of the reference image and the images to be registered, performing approximate nearest neighbor matching between the feature points of the reference image and the images to be registered by using a fast nearest neighbor search method, acquiring feature points in the images to be registered that are most similar to the feature points of the reference image, and generating a preliminary matched feature point pair; S224, based on the preliminary matched feature point pair, performing detection by using a random sample consensus method, eliminating incorrectly matched feature point pairs by constructing an affine transformation model, and generating an optimized feature point pair; S225, based on the optimized feature point pair, calculating an affine transformation matrix of the images to be registered to the reference image; S226, perform spatial transformation on each pixel of the to-be-registered image by using the affine transformation matrix to obtain new coordinates of the to-be-registered image in the reference image coordinate system, and finally generate a registered multi-source remote sensing image.

4. The forest resource prediction method based on remote sensing image analysis according to claim 1, characterized in that, Step S3 specifically comprises: S31, set a wavelet function and a decomposition layer number, and perform wavelet decomposition on each remote sensing image after registration to obtain low-frequency components and high-frequency components of each remote sensing image; S32, based on the low-frequency components, calculate the definition index of the low-frequency components of each remote sensing image, determine the weight of the low-frequency components, and generate a fused low-frequency component through weighted fusion, specifically: S321, calculate the average gradient of the low-frequency components of each remote sensing image, and take it as the definition index, that is: wherein, denotes the first denotes the average gradient of the low-frequency component of the remote sensing image, , denote the number of pixels of the low-frequency component along the horizontal and vertical directions, respectively, denotes the first denotes the gray value of the low-frequency component of the remote sensing image at pixel point , denotes the gray value of the low-frequency component of the remote sensing image at pixel point , denotes the gray value of the low-frequency component of the remote sensing image at pixel point , denotes the gray value of the low-frequency component of the remote sensing image at pixel point . S322, sort the average gradients of the low-frequency components of each remote sensing image from large to small, set a larger weight for a remote sensing image with a larger average gradient, and the sum of the weights of each remote sensing image is 1, to obtain the weight corresponding to the average gradient of the low-frequency component of each remote sensing image; S323, perform weighted averaging on the average gradient of the low-frequency component of each remote sensing image and the weight corresponding thereto to obtain the fused low-frequency component, that is: wherein, represents the low frequency component after fusion, represents the number of remote sensing images, represents the weight of the remote sensing image. S33, perform region energy-based fusion on the high-frequency components to generate a fused high-frequency component, specifically: S331, based on the high-frequency components, calculate the energy value of an n*n region around each pixel point of each remote sensing image, that is: wherein, represents the energy value of the high-frequency component of the remote sensing image in the n x n region around the pixel point , represents the high-frequency coefficient of the high-frequency component of the remote sensing image at the pixel point , represents the size of the region around the pixel point, represents a floor operation; S332, for each pixel point, compare the energy values of all remote sensing images at the pixel point, select the remote sensing image with the maximum energy value, take the high-frequency coefficient of the remote sensing image as the optimal high-frequency coefficient of the pixel point, and finally splice the optimal high-frequency coefficients of all pixel points to generate the fused high-frequency component, that is: wherein, represents the coefficient of the high-frequency component after fusion at the pixel point , represents the high-frequency coefficient of the remote sensing image when the energy value is maximum at the pixel point , represents the remote sensing image when the energy value is maximum at the pixel point , represents the argument obtained when the function takes the maximum value, i.e. the remote sensing image ; S34, perform inverse wavelet transformation on the fused low-frequency component and the high-frequency component to generate a fused remote sensing image.

5. The forest resource prediction method based on remote sensing image analysis according to claim 1, characterized in that, Step S4 specifically comprises: S41, based on the fused remote sensing image, extract the reflectivity values of the near-infrared band and the red band, calculate the normalized vegetation index value of each pixel point, and take it as the spectral feature, that is: wherein, represents a normalized vegetation index value of each pixel point, represents reflectance of the near-infrared band, represents reflectance of the red light band; S42, splice the normalized vegetation index value of each pixel point to generate a spectral feature matrix.

6. The forest resource prediction method based on remote sensing image analysis according to claim 1, characterized in that, The actual observation value of the forest resource includes a forest coverage.

7. The forest resource prediction method based on remote sensing image analysis according to claim 1, characterized in that, The formula for calculating the mutual information value between each feature variable and the prediction target variable in step S5 is: wherein denotes the mutual information value between the feature variable and the prediction target variable , denotes the feature variable, denotes the prediction target variable, denotes the joint probability distribution of the feature variable and the prediction target variable , denotes the logarithm function, denotes the marginal probability distribution of the feature variable , denotes the marginal probability distribution of the prediction target variable .

8. The forest resource prediction method based on remote sensing image analysis according to claim 1, characterized in that, Step S6 specifically comprises: S61, randomly sample the optimized spectral features to construct multiple decision trees; S62, perform feature evaluation on each decision tree to construct improved multiple decision trees, specifically: S621, for each node of each decision tree, calculate the weighted Gini index and information gain of each candidate feature; S622, assign weights to the weighted Gini index and information gain by using the analytic hierarchy process method to calculate the comprehensive value of each candidate feature; S623, select the candidate feature with the optimal comprehensive value as the split feature to realize improvement of a single decision tree, and repeatedly execute S621-S623 to realize improvement of multiple decision trees, and finally generate improved multiple decision trees; S63, group the improved multiple decision trees into a random forest model to generate an improved random forest model.

9. The forest resource prediction method based on remote sensing image analysis according to claim 8, characterized in that, The formula for calculating the weighted Gini index of each candidate feature and the information gain is: in, Indicates the current node The Gini index, This represents the total number of categories of samples in the current node. Indicates the first The proportion of class samples in the current node. Indicates the current node The sample set, Representing candidate features, Representing candidate features The number of possible values, Representing candidate features Values a subset of samples Represents a subset of samples The Gini index, Represents a subset of samples The Middle The proportion of class samples, Representing candidate features The weighted Gini index, Represents the sample set of the current node Information gain of candidate features Represents the sample set Information entropy Represents a subset of samples Information entropy This represents the logarithmic function with base 2.

10. The forest resource prediction method based on remote sensing image analysis according to claim 8, characterized in that, The formula for calculating the comprehensive value of each candidate feature is: wherein, denotes a combined value for each candidate feature, , denote a weighted Gini index and an information gain for a certain candidate feature, respectively, , denote weights for the weighted Gini index and the information gain, respectively.

Images