Deep learning-based euryale ferox traditional Chinese medicinal material remote sensing identification method

Abstract

The invention relates to the technical field of remote sensing image processing, in particular to a deep learning-based euryale ferox traditional Chinese medicinal material remote sensing recognition method, which comprises the steps of acquiring data, extracting features, constructing an optimal feature set, constructing a convolutional neural network model and training the model. According to the method, crops with high spectral similarity can be distinguished with high precision, the whole process is automatic, manual intervention is not needed in the phenological feature extraction, feature optimization and intelligent interpretation processes, subjective deviation caused by manual experience is avoided, and compared with a manual assistance method, the overall processing efficiency is also improved, and the method is suitable for popularization and application. The method can be quickly popularized to automatic interpretation in a large area; meanwhile, optical and radar multi-source data are compatible, high classification precision can still be kept under the condition that optical images are interfered by clouds and rains or crop plots are seriously broken, and the applicability in complex scenes is remarkably improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of remote sensing image processing, and particularly relates to a remote sensing identification method for Euryale ferox medicinal materials based on deep learning. BACKGROUND

[0002] Remote sensing interpretation refers to a process of modeling, analyzing, reasoning and judging the features of ground objects in remote sensing images according to actual application requirements, combining interpretation marks and interpretation experience, and finally obtaining ground object information. This technology has been widely applied in the field of crop identification and monitoring due to its advantages of macroscopicity, high efficiency and low cost. In traditional technology, common interpretation methods mainly include vegetation index and spectral feature extraction, such as using NDVI, EVI and other indexes to distinguish vegetation coverage and crop types. This kind of method is simple to operate, low in calculation cost, and has good interpretability, but usually relies on the experience of researchers for interpretation, and the degree of automation and intelligence is insufficient. For some crops with small planting scale and specific biological characteristics, traditional methods mainly rely on ground investigation and statistical methods to obtain distribution information, that is, through field sampling and regular investigation to obtain planting area and growth conditions. Although this kind of method has high precision, it is difficult to meet the needs of large-scale and real-time monitoring due to long investigation period, high labor cost and insufficient spatial coverage. In recent years, traditional machine learning classifiers such as decision tree, random forest and support vector machine have been widely applied in crop identification, which can improve the classification accuracy to a certain extent, but their ability to distinguish crops with high spectral similarity is still limited. In order to make up for the shortcomings of the above methods, in recent years, machine learning methods have been gradually introduced into the field of remote sensing crop identification. Different classifiers such as decision tree, random forest and support vector machine can improve the classification accuracy to a certain extent, especially in the multi-dimensional feature environment, which shows strong stability and applicability.

[0003] The existing remote sensing interpretation techniques mainly include three parts: data selection, feature selection, and classifier selection. In terms of data selection, optical images are the most commonly used data source, and different land types are usually distinguished by extracting spectral bands and constructing vegetation indices. In contrast, synthetic aperture radar (SAR) data has unique advantages: long-wave microwaves can penetrate vegetation cover, are sensitive to wetlands and water bodies, and are not limited by climate and light conditions, so they are widely used in aquatic vegetation monitoring. Previous studies have shown that the fusion of optical and radar data can fully leverage their complementary advantages and improve classification accuracy. However, existing fusion methods mostly remain in a static superposition state, lacking a fusion strategy for crop phenological characteristics and time series dynamic changes, limiting the potential for further improving recognition accuracy. In terms of feature selection, the introduction of multi-dimensional features can provide more information for classification, but it also brings feature redundancy and increased model complexity. Common feature selection methods include: filter methods (such as ReliefF), which have the advantages of high computational efficiency and suitability for large-scale feature selection, but are independent of the subsequent classification process and cannot consider feature interactions; wrapper methods (such as recursive feature elimination), which use classification performance as the evaluation standard and can obtain a better feature set, but have a higher computational cost; embedded methods (such as feature importance ranking based on random forests), which can complete feature selection during training, balancing efficiency and effectiveness, but the results are easily affected by algorithm bias. In complex remote sensing scenarios, a single method often struggles to balance efficiency and accuracy. In terms of classifier selection, traditional machine learning methods can be divided into parameterized and non-parameterized methods. Parameterized methods (such as maximum likelihood) rely on data distribution assumptions and are suitable for single-modal data, but their performance is limited in multi-modal complex data. Non-parameterized methods (such as support vector machines and random forests) are more versatile as they do not rely on distribution assumptions. Among them, support vector machines perform well in small sample and multi-spectral conditions, but have limitations in large-scale and multi-class tasks; random forests can handle high-dimensional features and noisy data through ensemble and voting mechanisms, and can quantify feature importance, so they are widely used in remote sensing classification. In recent years, with the development of deep learning, convolutional neural networks (CNN) have been rapidly promoted in remote sensing classification tasks. CNN can automatically extract multi-level features of images through multiple convolution and pooling operations, effectively representing spatial patterns, and has obvious advantages in high-resolution and complex scenarios. However, CNN is highly dependent on large-scale labeled samples and computational resources, and has high model complexity, which still poses certain limitations in practical application.

[0004] In summary, the existing remote sensing interpretation technology has deficiencies in data utilization, feature optimization and classification algorithm. The fusion of optical and radar data does not fully consider the dynamic changes of crop phenology. The single feature selection method has a trade-off between efficiency and accuracy. The traditional classifier has limited ability to distinguish similar spectral crops. Although deep learning models have potential, they are resource and sample dependent. To address these issues, the present patent proposes an intelligent remote sensing interpretation method based on phenological feature driving. This method does not rely on manual experience for feature selection, can extract key phenological windows, and combine multiple feature optimization mechanisms to form an optimal feature set. Then, the feature set is input into a convolutional neural network classifier to achieve high-precision intelligent interpretation of crops. This method significantly improves the automation and intelligence level of large-scale crop remote sensing interpretation while ensuring classification accuracy. SUMMARY

[0005] The present application aims to provide a deep learning-based gualou seed traditional Chinese medicine remote sensing identification method to improve the problems in the prior art.

[0006] The present application is implemented as follows: a deep learning-based gualou seed traditional Chinese medicine remote sensing identification method includes the following steps:

[0007] Step 1: Obtain image data and pre-process the image data to obtain pre-processed data; the image data includes multi-temporal data of optical remote sensing image data and radar remote sensing image data; obtain ground sample data, including sample points and land class types; wherein the sample points include gualou seed sample points and easily confused land class sample points;

[0008] Step 2: Extract optical features and radar polarization features, group the obtained multi-temporal data by month, apply the median value synthesis method to all effective images in each month, and perform synthesis operation on the optical features and radar polarization features to obtain feature sequences in units of months, and construct a multi-dimensional time series feature set as an initial feature set;

[0009] Step 3: Analyze and compare the time series curves of each feature in the initial feature set, combine the specificity of gualou seed in the growth cycle and canopy succession, identify the time series stage with the most significant feature difference between gualou seed and easily confused land classes according to the relative change law of feature values, determine the key phenological window, and construct an optimized feature set based on the key phenological window;

[0010] Step 4: Construct a convolutional neural network model, wherein the convolutional neural network model includes an input layer, a feature extraction layer, a global average pooling layer and an output layer;

[0011] Input layer: input the optimized feature set as a multi-channel matrix of HxWxC, where H and W are spatial dimensions, and C is the number of channels, i.e. the number of features in the optimized feature set;

[0012] Feature extraction layer: extracted by three convolutional blocks, each of which includes a convolution operation, batch normalization and a ReLU activation function, the size of the convolution kernel is 3x3, and the number of channels is expanded from the input channel to 32, 64 and 128 in turn;

[0013] Global average pooling layer: after the convolution feature extraction is completed, the HxW feature matrix of each channel is reduced to 1x1 by global average pooling to generate a 128-dimensional feature vector;

[0014] Output layer: including two fully connected layers, the first fully connected layer is used to map the reduced feature vector to 64 dimensions, then the Dropout processing is performed, and finally the second fully connected layer is used to map to each category to output the classification result;

[0015] Step 5, convert the sample points and their neighborhood pixels into a two-dimensional array, and extract the feature matrix of the ground sample data based on the preferred feature set, then divide it into a training set and a validation set according to a ratio of 7:3, input the training set into the convolutional neural network model, and train the convolutional neural network model using a multi-class supervised learning method to obtain a trained convolutional neural network model; wherein, the training of the convolutional neural network model comprises:

[0016] Loss function setting: cross-entropy loss function is selected as the optimization objective;

[0017] Optimizer and parameter configuration: the initial learning rate is set to 0.0001-0.01, the training rounds are in the range of 50-200 rounds, and the regularization constraint is introduced in the parameter update;

[0018] Learning rate scheduling strategy: set the StepLR learning rate scheduler, and dynamically decay the learning rate according to the iteration rounds during the training process.

[0019] More preferably, in step 1, the easily confused classes include lotus and rice.

[0020] More preferably, in step 1, the pre-processing of the image data specifically includes:

[0021] For optical remote sensing image data, the steps of radiometric calibration, atmospheric correction, cloud and cirrus mask, geometric correction and re-projection are adopted; for radar remote sensing image data, speckle noise suppression method is used to reduce speckle interference while preserving texture and edge features, extract VV and VH two polarization channels, and construct derived features, the derived features include polarization ratio, difference.

[0022] More preferably, in step 2, the moon median synthesis method specifically includes:

[0023] Suppose there are n images within this time frame, and the value of a certain pixel in the i-th image is... ;

[0024] The value of this pixel in the median composite image is ;

[0025] The median(...) is a mathematical operation for finding the median, which means taking the middle number after sorting a set of numbers by size (if the number of numbers is even, then the average of the two middle numbers is taken). It is the value of a certain pixel in the first image, and so on.

[0026] More preferably, step 2, extracting optical features and radar polarization features, specifically includes:

[0027] The optical feature extraction involves calculating several vegetation indices, moisture indices, and red edge indices from the preprocessed optical remote sensing image data. The vegetation indices include NDVI, EVI, SAVI, RVI, DVI, EVI2, GRVI, GNDVI, PRSI, EVSI, and FAI. The moisture indices include LSWI, MDWI, MSI, and MNDWI. The red edge indices include NDVIre1-3 and Clre1-3.

[0028] The radar polarization feature extraction involves constructing features from the VV and VH polarization data in the preprocessed radar remote sensing image data. This includes directly using VV and VH polarization images, and calculating the total backscattering intensity and differential scattering intensity based on the VV and VH data, where the total backscattering intensity = VH + VV and the differential scattering intensity = VH - VV.

[0029] More preferably, in step 2, the determination of valid images is as follows: first, use cloud cover to filter and select images with cloud cover less than 10%; then, use Sentinel data QA60 band to filter cirrus clouds and select images corresponding to pixels marked as "non-cloud" and "non-cirrus" in this band.

[0030] The method for synthesizing the median value of the month involves spatially registering and superimposing multiple images within a certain time range. For each pixel at a given location, the median of all values ​​in the time series is taken. If the number of values ​​is even, the average of the two middle values ​​is taken.

[0031] More preferably, in step 3, constructing the preferred feature set based on the key phenological window specifically includes:

[0032] ① Preliminary feature screening: Extract the feature subset corresponding to the key phenological window from the initial feature set, remove redundant features outside the key phenological window, and obtain the key phenological window feature set;

[0033] ②ReliefF Feature Evaluation: Randomly sample from the feature set of the key object window, calculate the similarity of each feature among samples of the same class and the difference among samples of different classes, assign weights to each feature according to the calculation results, continuously update the weights of all features during the iteration process until all feature weights converge, sort according to weights, remove features with low contribution, and retain the top-ranked candidate features.

[0034] ③J-M distance feature assessment: For each feature in the key object window feature set, calculate the separability index between water chestnut and easily confused land cover. By comparing the separation degree of different category pairs, screen out features with insufficient distinguishing ability and retain features with high separation degree in the target category pairs.

[0035] ④ Construction of the optimal feature set: Combining the overall importance assessment results of ReliefF and the category distinction results of JM distance, cross-comparison is performed, and only features that perform well in both indicators are retained to form the optimal feature set.

[0036] More preferably, in the ReliefF feature evaluation, the feature weight update rule is as follows:

[0037] ;

[0038] Where j represents a certain feature parameter; is the weight of feature j, which is usually initialized to 0; m is the number of samples drawn. The weights of feature j; Let be the random sample selected in the i-th iteration; k_neighbors is the number of nearest neighbors selected. For the sample Nearest neighbor samples of the same category; For the sample Nearest neighbor samples of different categories; For the sample Category; The proportion of category C, For category The proportion, For the sample Nearest neighbor samples of the same type The difference in characteristic parameter j;

[0039] In the JM distance feature evaluation, the derivation process of the separability metric J is as follows:

[0040] First, calculate the distance B from Bhattacharyya:

[0041] ;

[0042] Then, based on B, derive the distance J between J and M:

[0043] ;

[0044] Where B represents the Bhattacharyya distance, used to measure the degree of overlap between two classes of data; a larger value indicates a more significant difference in distribution. , and The mean and variance of a specific feature representing two different categories; This transforms B into a normalized JM distance, with a value ranging from 0 to 2. A larger value indicates a greater separability between the two categories.

[0045] More preferably, in step 4, the calculation formulas for convolution operation and batch normalization are as follows:

[0046] ;

[0047] ;

[0048] in, This represents the output of the convolutional layer. This represents the result after batch normalization. For input features, and For the kernel weights and biases, It is the ReLU activation function. and Let be the mean and variance of the k-th channel in the mini-batch. and These are the trainable scaling and translation parameters in the batch normalization layer. To prevent extremely small constants with a denominator of zero;

[0049] The formula for calculating global average pooling is:

[0050] ;

[0051] in, H represents the average value of all pixels in the k-th channel feature map, where H and W are the height and width of the feature map. This represents the feature value of the k-th channel at position (i, j) in the feature map, and is also the output of the convolutional layer;

[0052] The calculation formula for Dropout processing is:

[0053]

[0054] in, This refers to the output result after Dropout processing. These are input features or activation values ​​of neurons. is a binary random variable that follows a Bernoulli distribution, representing whether to retain the neuron. The probability of each neuron being retained is 1-p, and here p is set to 0.5.

[0055] More preferably, in step 5, the formula for calculating the cross-entropy loss function is:

[0056] ;

[0057] Where C represents the total number of categories, This represents the one-hot label for category c. This represents the soft max probability output for category c. The formula for calculating the Soft max probability output is:

[0058] ;

[0059] Where 'c' represents a category index, and C represents the total number of categories. This represents the unnormalized score of the c-th class in the output layer of the neural network. This represents the indexed score for class c. This represents the summation of the indexed scores for all categories, used for normalization. This represents the probability that the input x belongs to class c. Indicates the final predicted category label;

[0060] In the learning rate scheduling strategy, the formula for calculating the exponential learning rate decay is:

[0061]

[0062] in, This represents the learning rate in the t-th iteration. This represents the initial learning rate. The decay coefficient is represented by s=10, which means that the decay occurs once every 10 training iterations, and t represents the number of iterations.

[0063] Compared with the prior art, the present invention has the following advantages:

[0064] 1. This method can distinguish crops with high spectral similarity with high accuracy.

[0065] 2. The entire process is automated, requiring no human intervention in phenological feature extraction, feature optimization, and intelligent interpretation. This avoids subjective biases caused by human experience and improves overall processing efficiency compared to manual assistance methods. It can be quickly extended to automated interpretation over a wide area.

[0066] 3. It is compatible with multi-source optical and radar data, and can maintain high classification accuracy even when optical images are affected by cloud and rain or crop plots are severely fragmented, significantly improving its applicability in complex scenarios.

[0067] 4. It relies solely on publicly available Sentinel series remote sensing data, eliminating the need for costly specialized equipment or large-scale, high-precision labeled data. The methodology can be encapsulated into a Python toolkit (based on the TensorFlow / PyTorch framework), allowing for direct application and a low barrier to entry, making it highly valuable for widespread adoption. Attached Figure Description

[0068] Figure 1 This is the ReliefF feature contribution evaluation diagram of the present invention.

[0069] Figure 2 This is the JM distance heat map of the present invention.

[0070] Figure 3 This is a comparison chart of the accuracy of key phenological periods and non-key phenological periods in this invention.

[0071] Figure 4 This is a comparison chart of the accuracy of different classification models of this invention.

[0072] Figure 5 This is an RGB remote sensing image of the study area of ​​this invention.

[0073] Figure 6 This is a comparison chart of image details predicted by different classification models of the present invention.

[0074] Figure 7 This is a technical flowchart of the present invention. Detailed Implementation

[0075] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below. Apparently, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention. Unless otherwise defined, the technical terms or scientific terms used herein shall have the ordinary meanings as understood by those of ordinary skill in the art to which the present invention pertains. The words such as "including" used herein mean that the elements or objects appearing before this word cover the elements or objects listed after this word and their equivalents, without excluding other elements or objects.

[0076] Embodiment 1: This embodiment provides a remote sensing identification method for Euryale ferox Chinese medicinal materials based on deep learning. As shown in Figures 1-7 the following, it includes the following steps:

[0077] Step 1: Obtain image data and preprocess the image data to obtain preprocessed data;

[0078] Specifically, Step 1 includes: obtaining multi-temporal data of optical remote sensing image data and radar remote sensing image data of the target area through the GEE (Google Earth Engine) platform; obtaining ground sample data, including sample points and land type; specifically, more than 2,000 sample points of Euryale ferox and easily confused crops (such as lotus, rice, etc.) are collected on the ground, covering different land types in the study area, covering categories such as Euryale ferox, lotus, rice, eel breeding areas, water areas, other vegetation, and construction land;

[0079] For the optical remote sensing image data, perform steps of radiometric calibration, atmospheric correction, cloud and cirrus masking, geometric correction, and reprojection to obtain preprocessed optical remote sensing image data; radiometric calibration is used to convert digital numbers (DN) into radiance or apparent reflectance; atmospheric correction corrects the reflectance data based on the atmospheric radiation transfer model to eliminate the influence of aerosols and water vapor and obtain the surface reflectance; cloud and cirrus masking is used to剔除受云干扰的区域; geometric correction and reprojection ensure the precise alignment of different images in space and project them to the same coordinate system uniformly;

[0080] For the radar remote sensing image data, use a speckle noise suppression method to reduce speckle interference while retaining texture and edge features, extract two polarization channels of VV and VH, and construct derived features, including polarization ratio and difference, to enhance the discrimination ability between land types and obtain preprocessed radar remote sensing image data.

[0081] Step 2: Extract optical features and radar polarization features. Then, group the acquired multi-temporal data by month. Apply the monthly median synthesis method to all valid images in each month to synthesize optical features and radar polarization features to obtain a feature sequence in units of months, so that each feature has a unique numerical expression in each month, thereby constructing a complete multi-dimensional time series feature set as the initial feature set.

[0082] Specifically, the extraction of optical features and radar polarization features includes:

[0083] ① Optical feature extraction specifically involves calculating several vegetation indices, moisture indices, and red edge indices from preprocessed optical remote sensing image data. These include 11 vegetation indices (such as NDVI, EVI, SAVI, RVI, DVI, EVI2, GRVI, GNDVI, PRSI, EVSI, and FAI), 4 moisture indices (such as LSWI, MDWI, MSI, and MNDWI), and 2 red edge indices (such as NDVIre1-3 and Clre1-3). Through the calculation of these indices, multidimensional feature information of the target area in terms of chlorophyll content, canopy structure, and water content can be obtained to support subsequent classification and identification steps.

[0084] ② Radar polarization feature extraction is specifically manifested as follows: feature construction is performed on the VV and VH polarization data in the preprocessed radar remote sensing image data. In addition to directly using the VV and VH polarization images, derived parameters such as total backscattering intensity and differential scattering intensity are further calculated based on the VV and VH data. Among them, total backscattering intensity = VH + VV, and differential scattering intensity = VH - VV. Through this processing step, the differential features of vegetation in terms of vertical structure, horizontal scattering, and water content can be extracted to enhance the ability to distinguish target vegetation categories.

[0085] Furthermore, the determination of valid images is as follows: first, cloud cover is used for filtering, and images with cloud cover of less than 10% are selected as valid. Then, cirrus cloud filtering is performed. Using the Sentinel data QA60 band, only pixels marked as "non-cloud" and "non-cirrus" in this band are selected as valid images.

[0086] Furthermore, the median composite method spatially registers and overlays multiple images within a certain time range. For each pixel at a given location, the median of all values ​​in the time series is taken. If the number of values ​​is even, the average of the two middle values ​​is taken, ultimately resulting in a median composite image. Specifically, this includes:

[0087] Suppose there are n images within this time frame, and the value of a certain pixel in the i-th image is... ;

[0088] The value of this pixel in the median composite image is ;

[0089] The median(...) is a mathematical operation for finding the median, which means taking the middle number after sorting a set of numbers by size (if the number of numbers is even, then the average of the two middle numbers is taken). It is the value of a certain pixel in the first image, and so on;

[0090] This method ensures the robustness of the image and prevents it from being strongly affected by extreme values. It also has typicality and applicability and is often used for remote sensing data processing in areas with unstable atmospheric conditions and frequent cloud and rain. This formula can reduce the impact of extreme values ​​on temporal characteristics to a certain extent.

[0091] Step 3: Analyze and compare the time series curves of each feature in the initial feature set. Combining the specificity of water chestnut in the growth cycle and canopy succession, and based on the relative change law of feature values, identify the time series stage where the feature difference between water chestnut and easily confused land cover (including lotus, rice, etc.) is most significant, determine it as the key phenological window, and construct an optimal feature set based on the key phenological window.

[0092] Specifically, in the implementation process, this embodiment addresses the unique characteristics of the growth cycle and canopy succession of *Euryale ferox* by synthesizing monthly median values ​​from multi-temporal optical images and radar polarimetric images to extract temporal features of vegetation indices and polarization characteristics. Subsequently, the temporal curves of each feature in the initial feature set are analyzed and compared, and key phenological windows are determined based on the relative change patterns of feature values. These windows can serve as the basis for feature optimization and classification model input. Through the above method, key temporal stages can be identified based on temporal features without human experience, thus providing stable feature support for the high-precision classification of *Euryale ferox*.

[0093] Specifically, after constructing the initial feature set and determining the key phenological windows, a multi-stage feature optimization mechanism is used to screen out representative high-discretionary features of water chestnuts compared to other crops and land types, in order to reduce the interference of redundant information and noise on the high-precision identification of water chestnuts. In this embodiment, the ReliefF algorithm and the JM distance method are used to evaluate the features. The former is used to measure the importance of the feature in the overall sample, and the latter is used to quantify the separability between water chestnuts and easily confused land types. The overall importance evaluation results of ReliefF and the category discrimination results of JM distance are combined and cross-compared to retain only the features that perform well in both indicators, and finally form the most representative optimized feature set. This feature set has good stability and discriminative power, providing efficient input for the training of subsequent classification models.

[0094] Furthermore, constructing an optimal feature set based on key phenological windows specifically includes:

[0095] ① Preliminary feature screening: Extract the feature subset corresponding to the key phenological window from the initial feature set, remove redundant features outside the key phenological window, and obtain the feature set of the key phenological window;

[0096] ②ReliefF Feature Evaluation: Randomly sample from the key phenological window feature set, calculate the similarity of each feature among samples of the same class and the difference among samples of different classes, assign weights to each feature according to the calculation results, continuously update the weights of all features during the iteration process until all feature weights converge, sort according to weights, remove features with low contribution, and retain the top-ranked candidate features.

[0097] ③J-M distance feature assessment: For each feature in the key phenological window feature set, calculate the separability index between water chestnut and easily confused land cover. By comparing the separation degree of different category pairs, screen out features with insufficient distinguishing ability and retain features with high separation degree in the target category pairs.

[0098] ④ Construction of the optimal feature set: Combining the overall importance assessment results of ReliefF and the category distinction results of JM distance, cross-comparison is performed, and only features that perform well in both indicators are retained to form the optimal feature set.

[0099] Furthermore, the ReliefF feature evaluation specifically involves: randomly selecting a sample j from the critical period feature subset, and then calculating the dissimilar nearest neighbors (nearest neighbors from different classes of sample j) and the same-class nearest neighbors (nearest neighbors from the same class of sample j). After calculating the dissimilar nearest neighbors and the same-class nearest neighbors, the feature weight update rule is based on the differences between the samples in the feature dimensions:

[0100] ;

[0101] Where j represents a certain feature parameter; is the weight of feature j, which is usually initialized to 0; m is the number of samples drawn. The weights of feature j; Let be the random sample selected in the i-th iteration; k_neighbors is the number of nearest neighbors selected. For the sample Nearest neighbor samples of the same category; For the sample Nearest neighbor samples of different categories; For the sample Category; The proportion of category C, For category The proportion, For the sample Nearest neighbor samples of the same type The difference in characteristic parameter j;

[0102] In the JM distance feature assessment, the Bhattacharyya distance B is calculated using the mean and variance of the features of the two types of land features. Then, the separability measure J is derived based on B. The derivation process of the separability measure J is as follows:

[0103] First, calculate the distance B from Bhattacharyya.

[0104] ;

[0105] Then, based on B, the distance J between J and M is derived.

[0106] ;

[0107] Where B represents the Bhattacharyya distance, used to measure the degree of overlap between two classes of data; a larger value indicates a more significant difference in distribution. , and The mean and variance of a specific feature representing two different categories; This transforms B into a normalized JM distance, with a value ranging from 0 to 2. A larger value indicates a greater separability between the two categories.

[0108] Step 4: Construct a convolutional neural network model, which includes an input layer, a feature extraction layer, a global average pooling layer, and an output layer.

[0109] Specifically, the input layer will input the selected feature set as a multi-channel matrix of H × W × C, where H and W are the spatial dimensions and C is the number of channels, i.e. the number of features in the feature set; this step defines the input form of the model and realizes the transition from the feature construction stage to the model training stage.

[0110] The feature extraction layer extracts features through three convolutional blocks. Each convolutional block includes a convolution operation, batch normalization (BN), and a ReLU activation function. The kernel size is 3×3, and the number of channels expands sequentially from the input channels to 32, 64, and 128. The convolutional outputs are stabilized through batch normalization training, and the ReLU activation function enhances the non-linear expressive power of the network. This structure is used to progressively extract features from low to high levels, forming the core extraction part of the model.

[0111] After the convolutional features are extracted, the global average pooling layer reduces the dimension of the H×W feature matrix of each channel to 1×1, generating a 128-dimensional feature vector.

[0112] The output layer consists of two fully connected layers. First, the dimensionality-reduced feature vectors are mapped to 64 dimensions through the first fully connected layer, and then processed by Dropout to improve generalization ability. Finally, the second fully connected layer is used to map to each category, outputting the classification results of various landforms in the study area, thus achieving high-precision identification of water chestnut.

[0113] Furthermore, the formulas for convolution operations and batch normalization are as follows:

[0114] ;

[0115] ;

[0116] in, This represents the output of the convolutional layer. This represents the result after batch normalization. For input features, and For the kernel weights and biases, It is the ReLU activation function. and Let be the mean and variance of the k-th channel in the mini-batch. and These are the trainable scaling and translation parameters in the batch normalization layer. To prevent extremely small constants with a denominator of zero;

[0117] The formula for calculating Global Average Pooling (GAP) is as follows:

[0118] ;

[0119] in, H represents the average value of all pixels in the k-th channel feature map, where H and W are the height and width of the feature map. This represents the feature value of the k-th channel at position (i, j) in the feature map.

[0120] The calculation formula for Dropout processing (where p is the probability of Dropout):

[0121]

[0122] in, This refers to the output result after Dropout processing. These are input features or activation values ​​of neurons. It is a binary random variable that follows a Bernoulli distribution, representing whether to retain the neuron. The probability of each neuron being retained is 1-p, and here p is set to 0.5.

[0123] Step 5: Convert the sample point and its neighboring pixels (to enhance spatial information, a 3×3 area around the sample point is taken to form the neighborhood of a sample point) into a two-dimensional array, and extract the feature matrix of the ground sample data based on the optimized feature set. Then, divide it into a training set and a validation set according to a 7:3 ratio. Input the training set into the convolutional neural network model, and train the convolutional neural network model using a multi-class supervised learning method to obtain a trained convolutional neural network model. The training of the convolutional neural network model includes:

[0124] Loss function setting: Cross-entropy loss function is selected as the optimization objective to measure the difference between the predicted result and the true label. This loss function can effectively adapt to multi-class classification scenarios and ensure the model's ability to distinguish each class.

[0125] Optimizer and parameter configuration: The initial learning rate can be set to 0.0001–0.01, and the number of training rounds can be in the range of 50–200 rounds, depending on the model convergence. To avoid overfitting during training, regularization constraints, such as weight decay, can be introduced in parameter updates to enable the model to maintain classification ability while having good generalization ability.

[0126] Learning rate scheduling strategy: To improve the stability and convergence speed of the training process, a StepLR learning rate scheduler is set up. During the training process, the learning rate is dynamically decayed according to the iteration rounds, thereby dynamically adjusting the learning rate to adapt to the optimization needs of different training stages.

[0127] It should be clarified that the validation set, without being used in training, is used to evaluate the model's generalization ability and monitor whether the model is overfitting.

[0128] Furthermore, the formula for calculating the cross-entropy loss function is as follows:

[0129] ;

[0130] Where C represents the total number of categories, This represents the one-hot label for category c. This represents the soft max probability output for category c. The formula for calculating the Soft max probability output is:

[0131] ;

[0132] Where c represents a category index and C represents the total number of categories. This represents the unnormalized score of the c-th class in the output layer of the neural network. This represents the indexed score for class c. This represents the summation of the indexed scores for all categories, used for normalization. This represents the probability that the input x belongs to class c. Indicates the final predicted category label;

[0133] In the learning rate scheduling strategy, the formula for calculating the exponential learning rate decay is:

[0134]

[0135] in, This represents the learning rate in the t-th iteration. This represents the initial learning rate. The decay coefficient is represented by s=10, which means that the decay occurs once every 10 training iterations, and t represents the number of iterations.

[0136] Example 2: Based on Example 1, this example takes Yugan County, Jiangxi Province as the study area and applies Example 1 in practice. This area is a major gorgon fruit producing area in China and has a typical water-wetland environment. Gorgon fruit is intercropped with crops such as lotus and rice, and needs to be distinguished by high-precision remote sensing identification technology.

[0137] 1. Data acquisition and preprocessing;

[0138] Data Acquisition: Optical remote sensing imagery data was obtained using Sentinel-2 multi-temporal images with a spatial resolution of 10 m, covering the entire study area; radar remote sensing imagery data was obtained using Sentinel-1 multi-temporal images with polarization channels including VV and VH; ground sample data was obtained by collecting more than 2,000 sample points of water chestnut and easily confused crops (lotus, rice, etc.) in the field, covering different land types in the study area, including water chestnut, lotus, rice, eel farming areas, water bodies, other vegetation, and construction land.

[0139] Data preprocessing: Optical remote sensing image data undergoes radiometric calibration, atmospheric correction, cloud and cirrus cloud masking, geometric correction, and reprojection; radar remote sensing image data is processed using speckle noise suppression to reduce speckle interference while preserving texture and edge features, extracting the VV and VH polarization channels, and constructing derived features, including polarization ratio and difference.

[0140] 2. Feature construction and temporal synthesis;

[0141] Optical and radar polarization features are extracted, and the acquired multi-temporal data are grouped by month. The monthly median synthesis method is applied to all effective images in each month to synthesize optical and radar polarization features, resulting in a feature sequence in units of month. A multi-dimensional temporal feature set is constructed as the initial feature set.

[0142] The optical feature extraction involves calculating several vegetation indices, moisture indices, and red-edge indices from the preprocessed optical remote sensing image data. The vegetation indices include NDVI, EVI, SAVI, RVI, DVI, EVI2, GRVI, GNDVI, PRSI, EVSI, and FAI. The moisture indices include LSWI, MDWI, MSI, and MNDWI. The red-edge indices include NDVIre1-3 and Clre1-3. The radar polarization feature extraction involves constructing features from the VV and VH polarization data in the preprocessed radar remote sensing image data. This includes directly using VV and VH polarization images, and calculating the total backscattering intensity and differential scattering intensity based on the VV and VH data. The total backscattering intensity = VH + VV, and the differential scattering intensity = VH - VV.

[0143] 3. Feature optimization;

[0144] The time-series curves of each feature in the initial feature set were analyzed and compared. Combining the specificity of gorgon fruit in the growth cycle and canopy succession, and based on the relative change law of feature values, the time-series stage with the most significant difference between gorgon fruit and easily confused land cover was identified as the key phenological window. Based on the key phenological window, an optimal feature set was constructed.

[0145] Specifically, constructing the preferred feature set based on key phenological windows includes:

[0146] ① Preliminary feature screening: Extract the feature subset corresponding to the key phenological window from the initial feature set, remove redundant features outside the key phenological window, and obtain the key phenological window feature set;

[0147] ②ReliefF Feature Evaluation: Randomly sample from the feature set of the key object window, calculate the similarity of each feature among samples of the same class and the difference among samples of different classes, assign weights to each feature according to the calculation results, continuously update the weights of all features during the iteration process until all feature weights converge, sort according to weights, remove features with low contribution, and retain the top-ranked candidate features.

[0148] ③J-M distance feature assessment: For each feature in the key object window feature set, calculate the separability index between water chestnut and easily confused land cover. By comparing the separation degree of different category pairs, screen out features with insufficient distinguishing ability and retain features with high separation degree in the target category pairs.

[0149] ④ Construction of the optimal feature set: Combining the overall importance assessment results of ReliefF and the category distinction results of JM distance, cross-comparison is performed, and only features that perform well in both indicators are retained to form the optimal feature set.

[0150] Specifically, the ReliefF algorithm is used to calculate the importance weight of features in the overall sample, and the JM distance is used to evaluate the separation degree between water chestnut and easily confused land types. Through cross-comparison and screening, the final optimal feature set is constructed, which effectively reduces redundant and noisy features.

[0151] First, in the section on feature optimization and key phenological period determination, Figure 1 This represents the feature contribution of ReliefF, demonstrating the contribution of different features in the classification task. Figure 1 As can be seen, the spectral index with the highest contribution to classification is LSWI, followed by GRVI, and DVI, FAI, EVI, EVI2, SAVI and RVI also show high contribution. The top 8 features in terms of contribution are selected as the feature set after ReliefF feature selection.

[0152] Figure 2 This paper presents the results of discrimination calculations between different land cover types based on the JM distance method, helping to assess the separability of features. Class 1 represents water chestnut, Class 2 represents lotus, Class 3 represents aquaculture areas, and Class 4 represents rice. Figure 2 As can be seen, LSWI, GRVI, DVI, EVI, EVI2, FAI, and SAVI still exhibit high discriminative power, indicating that these features are highly separable when distinguishing different land use types. Although RVI has a lower overall discriminative power, it shows high separability in Class 1 and Class 2, indicating that it still has a significant impact on the classification of these two classes. Therefore, considering classification contribution, feature complementarity, and computational efficiency, LSWI, GRVI, DVI, EVI, EVI2, FAI, SAVI, and RVI are ultimately selected as the preferred features to improve the accuracy of the model.

[0153] Figure 3 This demonstrates a comparison of accuracy under different phenological stages, by Figure 3 As can be seen, the accuracy indicators of the key phenological period are higher than those of the non-key phenological period, proving that the time sequence stage with the most significant differences in identification features can significantly improve classification accuracy as the key phenological window, supporting the rationality of "constructing an optimal feature set based on the key phenological window".

[0154] Secondly, in the section on classification model selection, Figure 4 : Indicates a comparison of the accuracy of different classification models, by Figure 4 As can be seen, the accuracy indicators of the CNN classification model are higher than those of the random forest model and the support vector machine model, proving that the CNN model is superior to the random forest model and the support vector machine model in classification ability, and further verifying the advantages of the CNN model in remote sensing crop classification.

[0155] Figure 5 This is an RGB remote sensing image of the study area, in which five detailed regions are circled as... Figure 6 The location identifier of the specific model part.

[0156] Figure 6 The imagery shows the predicted results of the study area under different classification models. In the predicted images: red represents water chestnuts, orange represents lotus flowers, cyan represents aquaculture areas, yellow represents rice paddies, blue represents water bodies, green represents other vegetation, and gray represents built-up land. Figure 6 In the image, (c), (g), (k), (o), and (s) represent the original images of details 1, 2, 3, 4, and 5, respectively. Figure 6 In the image, (d), (h), (l), (p), and (t) represent the CNN model predictions for details 1, 2, 3, 4, and 5, respectively. Figure 6 In the image, (e), (i), (m), (q), and (u) represent the random forest model predictions for details 1, 2, 3, 4, and 5, respectively. Figure 6 In the image, (f), (j), (n), (r), and (v) represent the SVM model prediction images for details 1, 2, 3, 4, and 5, respectively.

[0157] 4. Model building and training;

[0158] The sample points (referring to location pixels containing true class labels, obtained through field surveys; each sample point is essentially a single pixel) and their neighboring pixels (referring to a 3×3 area around each sample point to enhance spatial information, forming a neighborhood of that sample point) are converted into a two-dimensional array. A feature matrix is ​​then extracted from the ground sample data based on a preferred feature set. Subsequently, the data is divided into training and validation sets in a 7:3 ratio and input into a lightweight convolutional neural network (CNN) model. The training set contains the preferred features of each training set sample point during key phenological periods, used to train the model, optimize model parameters, and provide comprehensive and stable input information for the CNN. The validation set, without participating in training, is used to evaluate the model's generalization ability and monitor for overfitting.

[0159] The structure of a convolutional neural network model includes an input layer, a feature extraction layer, a global average pooling layer, and an output layer.

[0160] The training parameters are set as follows: loss function is cross-entropy, optimizer is Adam, training epochs are 100 (when the number of epochs is ≥80, the accuracy of the validation set tends to be stable with fluctuations ≤0.5%, and continued iteration will lead to overfitting), and the initial learning rate is set to 0.001 (based on the adaptation of the dimension and sample size of the optimized feature set to avoid model oscillation caused by an excessively high learning rate).

[0161] 5. Classification prediction;

[0162] The trained CNN model was used to perform pixel-by-pixel prediction on the images of the study area to generate images of water chestnut classification results.

[0163] To verify the effectiveness and accuracy of feature optimization and deep learning framework, experimental results are presented in three main stages: determination of key phenological periods, feature optimization, and classification model selection. This comprehensively evaluates the application effect of each stage in remote sensing interpretation of water chestnut.

[0164] 6. Experimental results;

[0165] Through the above experimental process, the following results were obtained in this embodiment: (1) The optical feature combination extracted based on key phenological periods (including LSWI, GRVI, EVI, DVI, FAI, EVI2, SAVI, RVI) showed significant advantages in distinguishing between water chestnut and lotus, and the integration of radar VV / VH polarization features and derived parameters can effectively improve the separability of classification.

[0166] (2) The overall classification accuracy of the convolutional neural network model on the validation set reached about 93%, and the Kappa coefficient was about 0.92, showing good stability and reliability.

[0167] (3) This method generates a detailed image of the distribution of gorgon fruit in the study area, realizing the accurate identification and monitoring of gorgon fruit planting areas, which can be directly applied to planting management and industrial planning.

[0168] While embodiments of the present invention have been described in detail above, it will be apparent to those skilled in the art that various modifications and variations can be made to these embodiments. However, it should be understood that such modifications and variations fall within the scope and spirit of the invention as set forth in the claims. Furthermore, the invention described herein may have other embodiments and can be implemented or carried out in various ways.

Claims

1. A deep learning-based remote sensing identification method for gordon euryale seed Chinese medicinal material, characterized by, Comprising the following steps: Step 1, obtaining image data, and pre-processing the image data to obtain pre-processed data; the image data includes multi-temporal data of optical remote sensing image data and radar remote sensing image data; obtaining ground sample data, including sample points and land class types; wherein the sample points include Euryale ferox sample points and easily confused land class sample points; Step 2, extracting optical features and radar polarization features, grouping the obtained multi-temporal data by month, applying the median synthesis method in each month to all effective images in the month, and synthesizing the optical features and radar polarization features to obtain a feature sequence in units of months, and constructing a multi-dimensional time sequence feature set as an initial feature set; Step 3, analyzing and comparing the time sequence curves of each feature in the initial feature set, combining the specificity of Euryale ferox in the growth cycle and the crown layer succession, identifying the time sequence stage with the most significant feature difference between Euryale ferox and easily confused land classes according to the relative change law of the feature value, and determining it as a key phenology window, and constructing an optimized feature set based on the key phenology window; Step 4, constructing a convolutional neural network model, wherein the convolutional neural network model includes an input layer, a feature extraction layer, a global average pooling layer, and an output layer; The input layer: input the optimized feature set as a multi-channel matrix of HxWxC, where H and W are spatial dimensions, and C is the number of channels, i.e. the number of features in the optimized feature set; The feature extraction layer: extracted by three convolutional blocks, each convolutional block includes convolution operation, batch normalization and ReLU activation function, the size of the convolution kernel is 3x3, and the number of channels is expanded from the input channel to 32, 64 and 128; The global average pooling layer: after the convolution feature extraction is completed, the HxW feature matrix of each channel is reduced to 1x1 by global average pooling to generate a 128-dimensional feature vector; The output layer: including two fully connected layers, first mapping the reduced feature vector to 64 dimensions through the first fully connected layer, then performing Dropout processing, and finally mapping to each class through the second fully connected layer to output the classification result; Step 5, converting the sample points and their neighborhood pixels into a two-dimensional array, and extracting the feature matrix of the ground sample data based on the optimized feature set, then dividing it into a training set and a validation set according to a 7:3 ratio, inputting the training set into the convolutional neural network model, and training the convolutional neural network model using a multi-class supervised learning method to obtain a trained convolutional neural network model; wherein the convolutional neural network model training includes: Loss function setting: selecting cross-entropy loss function as the optimization target; Optimizer and parameter configuration: the initial learning rate is set to 0.0001-0.01, the training rounds are in the range of 50-200 rounds, and regularization constraints are introduced in the parameter update; Learning rate scheduling strategy: setting StepLR learning rate scheduler, and dynamically decaying the learning rate during the training process according to the iteration rounds.

2. The deep learning-based remote sensing identification method of Goutou Chinese medicinal material according to claim 1, characterized in that, In step 1, the easily confused land classes include lotus and rice.

3. The deep learning-based remote sensing identification method for Goutberry Traditional Chinese Medicine according to claim 2, characterized in that, In step 1, the pre-processing of the image data specifically includes: For optical remote sensing image data, the steps of radiation calibration, atmospheric correction, cloud and cirrus mask, geometric correction and re-projection are adopted; for radar remote sensing image data, speckle noise suppression method is adopted to reduce speckle interference, while retaining texture and edge features, extract VV and VH two polarization channels, and construct derived features, the derived features include polarization ratio, difference value.

4. The deep learning-based remote sensing identification method of Goutou Chinese medicinal material according to claim 3, characterized in that, In the step 2, the moon median synthesis method specifically comprises: Assume that there are n scenes of images in the time range, and the value of a pixel in the i-th scene of images is ; The value of the pixel in the median composite image is ; where median( ) is the mathematical operation of "median", i.e. the value in the middle of a list of numbers sorted in ascending order (if the number of values is even, the average of the two middle values is taken); is the value of a certain pixel in the 1st scene image, and so on.

5. The deep learning-based remote sensing identification method of Goutou Chinese medicinal material according to claim 4, characterized in that, In the step 2, the extraction of optical features and radar polarization features specifically comprises: The optical feature extraction is to calculate a plurality of vegetation indices, moisture indices and red edge indices from the pre-processed optical remote sensing image data, wherein the vegetation indices include NDVI, EVI, SAVI, RVI, DVI, EVI2, GRVI, GNDVI, PRSI, EVSI, FAI, the moisture indices include LSWI, MDWI, MSI, MNDWI, and the red edge indices include NDVIre1-3, Clre1-3; The radar polarization feature extraction is to perform feature construction on the VV polarization and VH polarization data in the pre-processed radar remote sensing image data, including directly using the VV and VH polarization images, and calculating the total backscatter intensity and differential scattering intensity based on the VV and VH data, wherein the total backscatter intensity = VH + VV, and the differential scattering intensity = VH - VV.

6. The deep learning-based remote sensing identification method of Goutou traditional Chinese medicinal material according to claim 5, characterized in that, In the step 2, the determination of the effective image is: first, filter using the cloud amount, select the image with a cloud amount less than 10%, and then filter using the Sentinel data QA60 band for cirrus filtering, select the image corresponding to the pixels marked as "non-cloud" and "non-cirrus" in the band; The moon median synthesis method is to register and superimpose multiple images in space within a certain time range, and take the median of all values in time sequence for each pixel, if the number of values is even, take the average of the middle two numbers.

7. The deep learning-based remote sensing recognition method of Goutou traditional Chinese medicinal material according to claim 6, characterized in that, In the step 3, the construction of the preferred feature set based on the key phenology window specifically comprises: ① Feature preliminary screening: extracting a feature subset corresponding to the key phenology window from the initial feature set, eliminating redundant features outside the key phenology window, and obtaining a key phenology window feature set; ② ReliefF feature evaluation: randomly extracting samples from the key phenology window feature set, calculating the similarity between the same samples and the difference between the different samples for each feature, assigning a weight to each feature according to the calculation result, updating the weight of all features to convergence in the iteration process, removing the features with low contribution according to the weight order, and retaining the candidate features ranked at the top; ③ J-M distance feature evaluation: for each feature in the key phenology window feature set, calculate the separability index of the target class and the easily confused class, compare the separation degrees of different class pairs, screen out features with insufficient separation ability, and retain features with high separation degree in the target class pair; ④ Preferred feature set construction: cross comparison combined with the overall importance evaluation result of ReliefF and the class separation result of J-M distance, only retain features that perform outstanding in both indicators, and form a preferred feature set.

8. The deep learning-based remote sensing recognition method of Goutou traditional Chinese medicinal material according to claim 7, characterized in that, In the ReliefF feature evaluation, the feature weight update rule is: ； wherein j represents a certain characteristic parameter; is the weight of the characteristic j, and its initial value is usually set to 0; m is the number of extracted samples, is the weight of the characteristic j; is the i-th selected random sample; k_neighbors is the number of selected neighbors; is the sample is the nearest neighbor sample of the same category; is the sample is the nearest neighbor sample of different categories; is the sample is the category of the sample is the proportion of the category C, is the category is the proportion of the category is the sample is the nearest neighbor sample of the same category is the difference in the characteristic parameter j; In the J-M distance feature evaluation, the separability measure J derivation process is: First, calculate the Bhattacharyya distance B: ； Then, derive J-M distance J based on B: ； where B represents the Bhattacharyya distance, which measures the degree of overlap of two classes of data, and the greater the value, the more significant the distribution difference, where 、 and represent the mean and variance of a specific feature of two different classes; is the conversion of B to the normalized J-M distance, which has a value range of 0 to 2, and the greater the value, the greater the separability between the two classes.

9. The deep learning-based remote sensing recognition method of Goutou traditional Chinese medicinal material according to claim 8, characterized in that, In step 4, the calculation formula of convolution operation and batch normalization is: ； ； wherein, represents the output of the convolutional layer, represents the result after batch normalization, is the input feature, and is the convolution kernel weight and bias, is the ReLU activation function, and is the mean and variance of the k-th channel in the mini-batch, and is the trainable scale and shift parameter in the batch normalization layer, is a very small constant to prevent the denominator from being zero; The calculation formula of global average pooling is: ； wherein, represents the average value of all pixels of the kth channel feature map, and H and W are the height and width of the feature map, represents the feature value of the kth channel at position (i, j) in the feature map, which is also the output of the convolution layer; The calculation formula of Dropout processing is: wherein, is the output result after Dropout processing, is the input feature or the activation value of the neuron, is a binary random variable subject to Bernoulli distribution, indicating whether to retain the neuron, and the probability of each neuron being retained is 1-p, where p is set to 0.

5.

10. The deep learning-based remote sensing recognition method of Goutonic Traditional Chinese Medicine according to claim 9, characterized in that, In step 5, the calculation formula of cross-entropy loss function is: ； where C denotes the total number of classes, one-hot label representing class c, Soft max probability output representing class c, The calculation formula of the Soft max probability output is: ； where c denotes a certain class index, C denotes the total number of classes, denotes the un-normalized score of the c-th class of the output layer of the neural network, denotes the exponentiated score of the c-th class, denotes the sum of the exponentiated scores over all classes for normalization, denotes the probability that the input x belongs to the c-th class, denotes the final predicted class label; In the learning rate scheduling strategy, the calculation formula of exponential learning rate decay is: wherein, denotes the learning rate for the tth iteration, denotes the initial learning rate, denotes the decay factor, s = 10 means decay once every 10 training, t denotes the iteration number.

Images