A Remote Sensing Identification Algorithm for Enteromorpha prolifera in the Yellow Sea Based on Edge Enhancement Deep Learning

Abstract

The present invention discloses a remote sensing recognition algorithm for Enteromorpha prolifera in the Yellow Sea based on edge-enhanced deep learning, which relates to the technical field of marine environmental information monitoring. The present invention proposes a remote sensing recognition method for Enteromorpha prolifera in the Yellow Sea based on edge-enhanced deep learning. By means of innovative measures such as constructing a specific model and screening key sample features, the recognition accuracy for fragmented Enteromorpha prolifera and its edges can be effectively improved, the generalization ability and robustness of the model can be enhanced, and the band limitation can be broken through, providing reliable technical support for the monitoring and ecological governance of Enteromorpha prolifera in the Yellow Sea.

Description

Technical Field

[0001] The present invention relates to the technical field of marine environmental information monitoring, and particularly relates to a remote sensing identification algorithm for Enteromorpha prolifera in the Yellow Sea based on edge-enhanced deep learning. Background Art

[0002] As the largest green tide disaster globally, the Enteromorpha prolifera green tide poses multiple threats to the marine ecosystem and economic society. Taking the outbreak of Enteromorpha prolifera in the Yellow Sea as an example, it not only damages aquaculture, blocks waterways, and threatens coastal tourism, but its decomposition also consumes dissolved oxygen in the water and releases hydrogen sulfide, leading to ecological imbalance in local waters.

[0003] Remote sensing monitoring of Enteromorpha prolifera, relying on large-scale and high-timeliness observation technologies that coordinate multiple platforms such as satellites and drones, can achieve precise identification of early small-scale Enteromorpha prolifera patches, dynamic inversion of the coverage area, and real-time tracking of the drift path, providing key data support for disaster warning, emergency response, and resource allocation. The long-term accumulated remote sensing data also reveals the correlation between the outbreak of Enteromorpha prolifera and climate change, eutrophication, and human activities, providing a scientific basis for ecological governance and policy optimization, and helping to explore the impact of the green tide on macro-ecological processes such as the carbon cycle and biodiversity, promoting the sustainable development of the marine environment.

[0004] In recent years, significant progress has been made in the remote sensing identification technology of the Enteromorpha prolifera green tide in the Yellow Sea, mainly relying on multi-source satellite data and algorithm innovation. Optical satellites play a dominant role, using the high reflectance characteristics of Enteromorpha prolifera in the near-infrared band to form a spectral difference with seawater for identification; synthetic aperture radar (SAR) data can be used as a supplement in bad weather.

[0005] Traditional remote sensing identification methods for Enteromorpha prolifera utilize the high reflectance characteristics of Enteromorpha prolifera in the near-infrared band and the spectral difference with seawater, and achieve detection through vegetation indices such as NDVI and FAI combined with threshold segmentation. However, it is easily affected by the environment and has high threshold sensitivity. Although various special indices have been developed subsequently to enhance the anti-interference ability, there are still deficiencies.

[0006] Currently, some studies have introduced artificial intelligence deep learning methods into the research of remote sensing identification of Enteromorpha prolifera to solve various defect problems encountered by traditional remote sensing identification methods for Enteromorpha prolifera. However, the inventor believes that in weak boundary scenarios such as fragmented Enteromorpha prolifera, the mixed edge area of seawater and Enteromorpha prolifera, etc., the deep learning network model still has problems of insufficient edge information extraction. The specific manifestations are as follows:

[0007] (1) The CNN model (Convolutional Neural Network) relies on data-driven learning of edge features and is prone to ignoring high-frequency details (such as weak boundaries) when the training samples are limited, resulting in blurred boundaries in the segmentation results.

[0008] (2) As the network deepens, the downsampling operation gradually attenuates high-order features such as edges, making it difficult for the decoding process to recover fine boundaries;

[0009] (3) The fusion of low-level features (edges) and high-level semantic information is easily affected by background noise, and the cross-scale feature alignment is insufficient, which affects the accuracy of boundary prediction.

[0010] Therefore, how to solve the above technical problems is a technical problem that technicians in this field urgently need to solve.

[0011] The information disclosed in this background technology section is only intended to enhance understanding of the overall background of the invention and should not be regarded as an admission or any form of suggestion that the information constitutes the prior art already known to a person skilled in the art. Summary of the Invention

[0012] In response to the above technical problems, an embodiment of the present invention provides a remote sensing identification algorithm for Enteromorpha in the Yellow Sea based on edge enhancement deep learning to solve the problems raised in the above background technology.

[0013] The present invention provides the following technical solution: a remote sensing recognition algorithm for Enteromorpha in the Yellow Sea based on edge-enhanced deep learning, comprising the following steps:

[0014] Build the EeaAglaeNet enteromorpha recognition model; the model structure includes: encoding layer, edge enhancement module and decoding layer;

[0015] Obtaining band and vegetation index characteristic parameters in the optical remote sensing image as initial sample data, and generating enteromorpha distribution image label data corresponding to the initial sample data;

[0016] After preprocessing, the initial sample data was input into the EeaAglaeNet enteromorpha recognition model for training. The SHAP analysis method was used to analyze the importance weight of the performance of the initial sample data in the model training to obtain the optimal sample data.

[0017] The optimal sample data include: Band 3 (0.560μm), Band 5 (0.705μm), Band 8 (0.842μm) and FAI; Band is the band, and FAI is the phytoplankton index;

[0018] The optimal sample data is input into the EeaAglaeNet Enteromorpha recognition model for training until the model is formed; the label data is the Enteromorpha distribution image and accuracy evaluation results;

[0019] The trained EeaAglaeNet enteromorpha recognition model is used to identify the distribution of enteromorpha.

[0020] Preferably, the encoding layer adopts a U-Net encoding layer structure, and the length, width and band of the input data are H, W and C respectively; the convolution operation uses a 3×3 convolution kernel, the activation function is ReLU, and the downsampling operation uses a 2×2 maximum pooling layer; after multiple convolutions, activations and downsampling, a feature map with 512 bands and a length and width of 1 / 8 of the original input length and width is obtained in the deepest layer.

[0021] Preferably, the input of the edge enhancement module includes three parts: first, the coding features after dimensionality reduction by 3×3 convolution from the coding layer; second, the hierarchical edge information extracted by the Laplacian pyramid; and third, the prediction results of the higher layers of the decoding layer downsampled to two dimensions.

[0022] These three sets of features are channel-joined and convolved to generate fused features; a spatial attention map is generated through a Sigmoid-activated convolutional layer, and a convolutional block attention module is used to calibrate the channel and spatial dimensions of the fused features, and finally output edge-enhanced features.

[0023] Preferably, the decoding layer adopts the U-Net upsampling path, and the input is the edge enhancement feature output by the edge feature enhancement module; the feature map resolution is improved layer by layer through transposed convolution or interpolation operations, the convolution operation adopts a 3×3 convolution kernel, the activation function is ReLU, and the upsampling operation uses a 2×2 maximum pooling layer; finally, the decoding feature maps of each layer are output to generate the enteromorpha recognition result corresponding to the input.

[0024] Preferably, the initial sample data includes: Band 1 (0.443 μm), Band 2 (0.490 μm), Band 3 (0.560 μm), Band 4 (0.665 μm), Band 5 (0.705 μm), Band 6 (0.740 μm), Band 7 (0.783 μm), Band 8 (0.842 μm), Band 8a (0.865 μm), Band 9 (0.945 μm), Band 11 (1.610 μm), and Band 12 (2.190 μm);

[0025] As well as NDVI, EVI, DVI, FAI and SRG; among them, NDVI is the Normalized Difference Vegetation Index, EVI is the Enhanced Vegetation Index, DVI is the Difference Vegetation Index, and SRG is the Simple Green Band Ratio Vegetation Index.

[0026] Preferably, the optimal sample data preprocessing method includes: radiometric calibration, atmospheric correction, resampling and land masking.

[0027] Preferably, the input feature dataset implements a three-stage segmentation strategy: First, the input features are randomly split into a training set and a test set at a ratio of 7:3. The test set is independently used to evaluate the generalization performance of the wind field reconstruction model. 20% of the samples are secondarily extracted from the training set to construct a validation set for real-time monitoring of the convergence stability during model training. Among them, the label data and the input feature dataset adopt the same partitioning process as the input feature data.

[0028] Preferably, the input feature dataset is standardized using the range normalization method.

[0029] Preferably, the loss function in the model adopts a composite loss function, which is jointly composed of binary cross-entropy and Dice loss.

[0030] Preferably, the accuracy evaluation result metrics are: Accuracy, Precision, Recall, F1 Score, and IoU;

[0031] Among them, Accuracy is the accuracy rate, Precision is the precision, Recall is the recall rate, F1 Score is the F1 score, and IoU is the intersection over union.

[0032] The remote sensing identification algorithm for Enteromorpha prolifera in the Yellow Sea based on edge-enhanced deep learning provided by the embodiments of the present invention has the following beneficial effects:

[0033] 1. The present invention introduces an edge enhancement module and designs a multi-scale pyramid structure to directly capture multi-scale edge information from the original image, making up for the blurred boundary of the segmentation result caused by the deficiency of CNN in weak boundary learning.

[0034] 2. Retain edge details at different resolutions to avoid edge degradation; integrate a lightweight convolutional block attention module to further optimize feature correlation and improve boundary localization accuracy.

[0035] 3. In addition, this method also has strong transferability and is not restricted by the band settings of remote sensing satellite data sources, thereby improving the generalization ability of the model and enhancing the robustness and accuracy of Enteromorpha prolifera remote sensing identification.

[0036] 4. When the present invention selects band data as input, it also introduces a variety of vegetation indices as physical information for auxiliary input; the vegetation index data can enhance the difference between Enteromorpha prolifera and other ground objects, highlight the characteristics of Enteromorpha prolifera, so as to more accurately judge the boundary position of Enteromorpha prolifera, reduce the recognition error caused by factors such as clouds, sea water, etc., and improve the accuracy and reliability of Enteromorpha prolifera edge recognition. Description of the Drawings

[0037] Figure 1It is the flow chart of the remote sensing recognition algorithm of Enteromorpha prolifera in the Yellow Sea based on edge-enhanced deep learning in the present invention;

[0038] Figure 2 It is the framework diagram of the Enteromorpha prolifera recognition network model EeaAglaeNet in the present invention;

[0039] Figure 3 It is the network framework diagram of the edge enhancement module in the present invention;

[0040] Figure 4 It is the result of the feature importance ranking in the present invention;

[0041] Figure 5 It is the comparison result of the EeaAglaeNet model and other classical Enteromorpha prolifera recognition model results. Specific embodiments

[0042] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative efforts belong to the scope of protection of the present invention.

[0043] In response to the problems mentioned in the above background technology, the embodiments of the present invention provide a remote sensing recognition algorithm of Enteromorpha prolifera in the Yellow Sea based on edge-enhanced deep learning to solve the above technical problems, and its technical solutions are as follows:

[0044] Next, in combination with the attached Figures 1-5 , and the specific embodiments will further illustrate the present invention.

[0045] The design idea and optimization steps of a remote sensing recognition algorithm of Enteromorpha prolifera in the Yellow Sea based on edge-enhanced deep learning provided by the present invention are as follows:

[0046] A remote sensing recognition method of Enteromorpha prolifera in the Yellow Sea based on edge-enhanced deep learning includes the following steps:

[0047] Step 1: Obtain remote sensing satellite data containing Enteromorpha prolifera; for optical remote sensing data, obtain the corresponding Enteromorpha prolifera distribution image data information.

[0048] Step 2: Preprocess the remote sensing satellite image data according to the relevant preprocessing methods of optical remote sensing images. The preprocessing process includes radiometric calibration, atmospheric correction, resampling, and land masking processing.

[0049] Step 3: Extract the band and vegetation index characteristic parameters in the optical remote sensing image; the vegetation index results and the data of each band together form a characteristic data set.

[0050] Among them, the vegetation indices are extracted using band calculations. The vegetation indices used are: NDVI, EVI, DVI, FAI, and SRG indices. The index calculation formulas are as follows:

[0051]

[0052] In the formula, NIR, RED, SWIR, GREEN, are the reflectance of the near-infrared band, the reflectance of the red band, the short-wave infrared reflectance, the reflectance of the green band, the wavelength of the near-infrared, the wavelength of the red, and the short-wave infrared wavelength in the optical image.

[0053] Step 4: Use the Python programming language to construct a remote sensing identification network for Enteromorpha prolifera in the Yellow Sea based on edge-enhanced deep learning of remote sensing optical datasets (EeaAglaeNet). The specific steps are as follows:

[0054] (1) Construction of the model input feature dataset and output features.

[0055] There are many band data involved in the imaging process of remote sensing optical data and current methods for monitoring Enteromorpha prolifera vegetation indices. The Enteromorpha prolifera identification environment is affected by clouds and seawater. To improve the accuracy of identifying the edge of Enteromorpha prolifera, in addition to selecting band data as input, this invention also introduces multiple vegetation indices as auxiliary inputs.

[0056] Taking Sentinel-2 as an example, the two input features contain a total of 17 features: band 1 (0.443μm), band 2 (0.490μm), band 3 (0.560μm), band 4 (0.665μm), band 5 (0.705μm), band 6 (0.740μm), band 7 (0.783μm), band 8 (0.842μm), band 8a (0.865μm), band 9 (0.945μm), band 11 (1.610μm), band 12 (2.190μm); as well as the Normalized Difference Vegetation Index (NDVI), Enhanced Vegetation Index (EVI), Difference Vegetation Index (DVI), and Floating Algae Index (FLA). The model uses the Green Band Ratio Vegetation Index (FAI) and the Simple Ratio Green (SRG) to construct an input feature dataset of the shape (n, x, y, c), where n is the number of samples, x and y are the length and width of the sample, respectively, and c is the number of features. The model's output features include an image of the Enteromorpha distribution and accuracy evaluation results.

[0057] (2) Feature importance analysis of the model.

[0058] Based on the SHAP interpretability model framework, we performed importance weight analysis on the multidimensional variables in the Enteromorpha feature dataset (n, x, y, c). By quantifying and ranking the contribution of each feature to the classification results, we selected the feature subset that drives the model's recognition performance, maximizing Enteromorpha detection accuracy. After screening, the optimal feature subset was found to be: band 3, band 5, band 8, and FAI.

[0059] (3) Division into training set, validation set and test set.

[0060] Based on the optimized feature dataset, a three-stage partitioning strategy was implemented. First, the input features were randomly split into training and test sets in a 7:3 ratio. The test set was independently used to evaluate the generalization performance of the Yellow Sea Enteromorpha remote sensing recognition model. Furthermore, a validation set was constructed by re-sampling 20% of the training set to monitor the convergence stability of the model during training. This approach also implemented a homogeneous partitioning process for the labeled data, ensuring strict spatiotemporal consistency of feature-label data pairs across training, validation, and testing.

[0061] (4) Data standardization (normalization)

[0062] Aiming at the modeling deviation problem caused by the dimensional differences of multi-dimensional features, the present invention constructs a data normalization preprocessing process. When the input parameters with different dimensions are not unified in scale, high-magnitude features are prone to dominate the model learning direction, while the contribution of low-magnitude features is nonlinearly compressed. To eliminate the interference of dimensional differences on the modeling process, this solution adopts the range normalization method, and compresses each eigenvalue to the interval [0, 1] through linear mapping. Its mathematical expression is: (6);

[0063] In the formula, and respectively represent the minimum observed value and the maximum observed value of the feature variable. While retaining the original data distribution law, this transformation realizes the balanced representation of each dimension feature and provides stable input for subsequent model construction.

[0064] (5) Model structure

[0065] The Enteromorpha prolifera recognition model EeaAglaeNet constructed by the present invention includes: an encoding layer (Encoder), an edge enhancement module, and a decoding layer (Decoder); among them, the network model framework of the Enteromorpha prolifera recognition network EeaAglaeNet is as Figure 2 shown.

[0066] First of all, the model retains the basic structure of the U-Net network, that is, it retains the structures of the encoding layer and the decoding layer. And it retains the method of introducing the data of the encoding layer during the decoding process to reduce data loss. Assume that the length, width, and number of bands of the input data are H, W, and C respectively. In the encoding module, the convolution operation uses a 3*3 convolution, the activation function is ReLu, and the downsampling operation uses a 2*2 max pooling layer. The deepest layer obtains a number of bands of 512, and the length and width are 1 / 8 of the original input length and width respectively.

[0067] In the edge enhancement module (Edge-Enhanced Attention, EEA), the main functions are: fusing the encoded features, high-frequency edge information, and high-level prediction features among multiple scales, and enhancing the ability to retain boundary details. The inputs of this module are: ① The encoded features after dimensionality reduction by 3×3 convolution from the encoding layer. ② The hierarchical edge information extracted by the Laplacian pyramid. ③ The prediction results downsampled to two dimensions from a higher level of the decoding layer. The above three groups of features are subjected to channel splicing and convolution operations to generate fused features. A spatial attention map (AttentionMASK) is generated through a convolution layer activated by Sigmoid to suppress noise and focus on key edge regions. The convolutional block attention module (CBAM) is used to calibrate the fused features in the channel and spatial dimensions, and the edge enhanced features are output. The network model framework of this module is as Figure 3 shown.

[0068] In the decoding module, a U-Net style upsampling path is adopted to gradually increase the resolution of the feature map layer by layer through transposed convolution or interpolation operations. The convolution operation uses a 3*3 convolution, the activation function is ReLu, and the upsampling operation uses a 2*2 max pooling layer. The input is the edge enhancement features output by the Edge Enhancement Module (EEA), and the output is the decoded feature maps at each level, finally generating the recognition result of the input Enteromorpha prolifera.

[0069] In terms of the design of the loss function, a composite loss function is adopted, which is jointly composed of Binary Cross-Entropy (BCE) and Dice loss. Among them, Binary Cross-Entropy is used to quantify the difference in the probability distribution between the predicted value and the true label, and Dice loss enhances the sensitivity of the model to the shape of the target area by calculating the proportion of the overlapping area between the predicted mask and the true mask. After weighted superposition of the two, the formula is expressed as:

[0070] (7);

[0071] In terms of the optimization strategy, the Adaptive Moment Estimation (Adam) algorithm is adopted, which combines the dynamic adjustment mechanisms of the first-order moment and the second-order moment to effectively balance the direction and step size of gradient update. The initial learning rate is set to 0.0001, and the parameter update amplitude is restricted by gradient clipping (ClipValue = 0.5) to further ensure the stability of the training process.

[0072] (4) Indexes for evaluating the accuracy of the model.

[0073] The indexes for evaluating the accuracy of the present invention are: Accuracy, Precision, Recall, F1 Score, Intersection over Union (IoU). Accuracy can measure the classification correct rate of the model on all samples, and the calculation formula is as follows: (8);

[0074] Precision can measure the proportion of actual positive classes among the positive samples predicted by the model, and the calculation formula is as follows:

[0075] (9);

[0076] Recall can measure the detection ability of the model for Enteromorpha prolifera targets, and the calculation formula is as follows:

[0077] (10);

[0078] The F1 score is the harmonic mean of precision and recall, and is calculated as follows:

[0079] (11);

[0080] The intersection-over-union (Io) can measure the degree of overlap between the Enteromorpha prediction results and the true labels, showing the degree of consistency between the Enteromorpha prediction results and the true situation. The calculation formula is as follows:

[0081] (12);

[0082] Among them, TP is a true positive example, TN is a true negative example, FP is a false positive example, and FN is a false negative example. Specific embodiments

[0083] This example uses multispectral imagery data from the Sentinel-2A / B satellites as a validation target to demonstrate a remote sensing monitoring method for Enteromorpha in the Yellow Sea based on an edge-enhanced deep learning framework. Users must install the Python 3.6.13 runtime environment on their local terminal and integrate the TensorFlow 1.15.0 and Keras 2.3.1 software libraries. TensorFlow is an open-source deep learning framework that supports distributed computing and provides low-level interfaces for building, training, and deploying neural network models. Keras is a high-level API based on TensorFlow. Its modular design simplifies the rapid construction and debugging of complex models.

[0084] Step 1: Input feature contribution analysis.

[0085] During model training, the number and combination of input features directly impacts training efficiency and predictive performance. Too few parameters can easily lead to underfitting, making it difficult for the model to capture underlying data correlations. Excessive parameters can increase the risk of overfitting, while also increasing computational complexity and inference latency. To optimize the input configuration of the EeaAglaeNet model, this solution uses SHAP (Shapley Additive Explanations) to evaluate the contribution of multi-source features, screening for feature sets with significant positive or negative impact on model output.

[0086] There are two major categories of input features in the SHAP method, with a total of 17 types. The first category is multispectral band feature data: band 1 (0.443μm), band 2 (0.490μm), band 3 (0.560μm), band 4 (0.665μm), band 5 (0.705μm), band 6 (0.740μm), band 7 (0.783μm), band 8 (0.842μm), band 8a (0.865μm), band 9 (0.945μm), band 11 (1.610μm) and band 12 (2.190μm). The second category is vegetation index feature data: NDVI, EVI, DVI, FAI and SRG vegetation index. The contribution of input features is as follows: Figure 4 shown.

[0087] The above input features were integrated to construct an optimized feature dataset with a data dimension of (n, 128, 128, 17). The dataset was divided according to a preset ratio: 70% of the total samples were used for model training (20% of which was used as a validation set for hyperparameter tuning), and the remaining 30% was used as an independent test set to ensure objectivity in model performance evaluation. The performance of the model in the test set was evaluated based on the quantitative evaluation results of feature contributions. Key spectral band and vegetation index combinations were selected, including Band 3 (0.560μm), Band 5 (0.705μm), Band 8 (0.842μm), and FAI. The above features were integrated to construct an optimized feature dataset.

[0088] Step 2: Accuracy Analysis. The optimized feature dataset selected in Step 1 was input into the Edge-Enhanced Deep Learning-Based Yellow Sea Enteromorpha Remote Sensing Identification Model (EeaAglaeNet) presented in this paper. The resulting accuracy evaluation index was 0.9762, and the precision reached 0.9530. By improving the loss function and enhancing the edge extraction module, the model's accuracy was significantly improved.

[0089] Step 3: Comparison with other classic Enteromorpha recognition models. To verify the effectiveness of EeaAglaeNet in remote sensing identification of Enteromorpha margins and sparse Enteromorpha in the Yellow Sea, the inversion results were compared with those of the classic neural network model U-Net and the lightweight Enteromorpha recognition model AglaeNet. The comparison results on the test set are shown in Table 1. Compared with other classic inversion models, the proposed EeaAglaeNet model improved its accuracy by 4.02%.

[0090] like Figure 5As shown in the figure, the model proposed by the present invention has a significant improvement in the extraction effect of the edges of Enteromorpha prolifera and fragmented Enteromorpha prolifera. By comparing the results of different models, it can be known that the U-Net deep learning network framework has good recognition ability in the field of Enteromorpha prolifera recognition in the Yellow Sea. The traditional U-Net deep learning model can achieve an overall recognition accuracy of 0.9128. However, in the lightweight deep learning network model AlgaeNet, since its design is a deep learning model for the red, green, and blue three bands, its performance on the multi-feature data set of remote sensing reflectance and vegetation index needs to be improved.

[0091] The EeaAglaeNet model proposed by the present invention can more effectively identify fragmented Enteromorpha prolifera and the edges of Enteromorpha prolifera, reduce the missed detection rate of Enteromorpha prolifera, and improve the accuracy and precision of Enteromorpha prolifera recognition. Figure 5 As shown in the figure, the results of Enteromorpha prolifera recognized by each model have been circled and marked with red boxes in the figure; from the circled results, it can be seen that the fragmented Enteromorpha prolifera and the edges of Enteromorpha prolifera that were not recognized by the traditional model method can have good recognition effects in the EeaAglaeNet model.

[0092] Table 1 Summary of model accuracy evaluation

[0093] Precision IoU Recall F1 Score U-Net 0.9128 0.7785 0.7785 0.8301 AglaeNet 0.8565 0.6582 0.7399 0.7940 EeaAglaeNet 0.9530 0.7016 0.7268 0.8246

[0094] In summary, the present invention proposes a method for remote sensing recognition of Enteromorpha prolifera in the Yellow Sea based on the fusion of edge enhancement and multi-scale attention, which can more effectively identify fragmented Enteromorpha prolifera in the Yellow Sea; the present invention has significant advantages in edge extraction, multi-scale fusion, and data adaptability; by introducing an edge enhancement mechanism and a multi-scale pyramid structure, the recognition accuracy of the model for fragmented Enteromorpha prolifera and its boundaries is effectively improved; combined with the channel-spatial attention mechanism, the correlation between features is enhanced, background noise interference is suppressed, and the recognition effect under complex sea conditions is improved.

[0095] At the same time, based on the analysis of feature importance, the present invention constructs a band-independent data input scheme, breaks through the dependence on specific remote sensing bands, and realizes the compatible migration of multi-source remote sensing data (such as optical, SAR); aiming at the characteristics of Enteromorpha prolifera in the Yellow Sea, a new loss function is designed to further improve the recognition accuracy of the edge region and the robustness of the model, realizing the fine extraction of fragmented Enteromorpha prolifera and the edges of Enteromorpha prolifera, and providing high-reliability technical support for disaster emergency response and ecological governance decision-making.

[0096] Although the specific implementation manners of the present invention are described above in conjunction with the accompanying drawings, it is not a limitation to the protection scope of the present invention. Those skilled in the art should understand that on the basis of the technical solutions of the present invention, various modifications or deformations that can be made by those skilled in the art without creative efforts are still within the protection scope of the present invention.

Claims

1. A remote sensing recognition algorithm for Enteromorpha prolifera in the Yellow Sea based on edge-enhanced deep learning, characterized in that, It includes the following steps: Build an Enteromorpha prolifera recognition model of EeaAglaeNet; the model structure includes: an encoding layer, an edge enhancement module, and a decoding layer; Obtain the band and vegetation index characteristic parameters in the optical remote sensing image as the initial sample data, and make the Enteromorpha prolifera distribution image label data corresponding to the initial sample data; Preprocess the initial sample data and input it into the Enteromorpha prolifera recognition model of EeaAglaeNet for training, and use the SHAP analysis method to analyze the importance weights of the performance of the initial sample data in the model training to obtain the optimal sample data; The optimal sample data includes: Band 3 - 0.560μm, Band 5 - 0.705μm, Band 8 - 0.842μm, and FAI; Band is the band, and FAI is the phytoplankton index; Input the optimal sample data into the Enteromorpha prolifera recognition model of EeaAglaeNet for training until the model is formed; the label data is the Enteromorpha prolifera distribution image and the accuracy evaluation result; Use the trained Enteromorpha prolifera recognition model of EeaAglaeNet to identify the Enteromorpha prolifera distribution; Among them, the encoding layer adopts the U-Net encoding layer structure, and the length, width, and band of the input data are H, W, and C respectively; the convolution operation uses a 3×3 convolution kernel, the activation function is ReLU, and the downsampling operation uses a 2×2 max pooling layer; after multiple convolutions, activations, and downsamplings, a feature map with a band number of 512 and a length and width that are 1 / 8 of the original input length and width is obtained at the deepest layer; The input of the edge enhancement module includes three parts. One is the encoded feature after dimensionality reduction by 3×3 convolution from the encoding layer; the second is the hierarchical edge information extracted by the Laplacian pyramid; the third is the prediction result downsampled to two dimensions from a higher level of the decoding layer; Perform channel splicing and convolution operations on these three groups of features to generate fused features; generate a spatial attention map through a convolution layer activated by Sigmoid, and use the convolutional block attention module to calibrate the fused features in the channel and spatial dimensions, and finally output the edge enhancement features; The decoding layer adopts the U-Net upsampling path, and the input is the edge enhancement features output by the edge feature enhancement module; the resolution of the feature map is gradually increased layer by layer through transposed convolution or interpolation operations. The convolution operation uses a 3×3 convolution kernel, the activation function is ReLU, and the upsampling operation uses a 2×2 max pooling layer; finally, the decoding feature maps of each level are output to generate the Enteromorpha prolifera recognition result corresponding to the input.

2. The remote sensing identification algorithm for Enteromorpha prolifera in the Yellow Sea based on edge-enhanced deep learning according to claim 1, wherein The initial sample data includes: Band 1 - 0.443μm, Band 2 - 0.490μm, Band 3 - 0.560μm, Band 4 - 0.665μm, Band 5 - 0.705μm, Band 6 - 0.740μm, Band 7 - 0.783μm, Band 8 - 0.842μm, Band 8a - 0.865μm, Band 9 - 0.945μm, Band 11 - 1.610μm, Band 12 - 2.190μm; And, NDVI, EVI, DVI, FAI, and SRG; where NDVI is the Normalized Difference Vegetation Index, EVI is the Enhanced Vegetation Index, DVI is the Difference Vegetation Index, and SRG is the Simple Ratio Green-band Vegetation Index.

3. The remote sensing identification algorithm for Enteromorpha prolifera in the Yellow Sea based on edge-enhanced deep learning according to claim 1, wherein The methods for optimal sample data preprocessing include: radiometric calibration, atmospheric correction, resampling, and land masking.

4. The remote sensing identification algorithm for Enteromorpha prolifera in the Yellow Sea based on edge-enhanced deep learning according to claim 1, wherein Implement a three-stage segmentation strategy for the input feature dataset: First, randomly split the input features into a training set and a test set at a ratio of 7:

3. The test set is independently used to evaluate the generalization performance of the remote sensing identification model for Enteromorpha prolifera in the Yellow Sea. Secondarily extract 20% of the samples from the training set to construct a validation set for real-time monitoring of the convergence stability during model training; among them, the labeled data and the input feature dataset adopt the same partitioning process as the input feature data.

5. The remote sensing recognition algorithm for Enteromorpha prolifera in the Yellow Sea based on edge-enhanced deep learning according to claim 1, characterized in that, The input feature dataset is standardized using the range normalization method.

6. The remote sensing identification algorithm for Enteromorpha prolifera in the Yellow Sea based on edge-enhanced deep learning according to claim 1, characterized in that, The loss function in the model adopts a composite loss function, which is jointly composed of binary cross-entropy and Dice loss.

7. The remote sensing identification algorithm for Enteromorpha prolifera in the Yellow Sea based on edge-enhanced deep learning according to claim 1, characterized in that, The accuracy evaluation result metrics are: Accuracy, Precision, Recall, F1 Score, and IoU. Among them, Accuracy is the accuracy rate, Precision is the precision, Recall is the recall rate, F1 Score is the F1 score, and IoU is the intersection over union.

Images