Multi-source remote sensing vegetation coverage inversion product fusion method, device and equipment

Abstract

This invention provides a method, apparatus, and device for fusing multi-source remote sensing vegetation cover inversion products, relating to the field of vegetation cover monitoring technology. The method includes: acquiring multi-source remote sensing vegetation cover inversion products; verifying the authenticity of the remote sensing vegetation cover inversion products to select the optimal remote sensing vegetation cover inversion product corresponding to each resolution; constructing a training dataset using the optimal remote sensing vegetation cover inversion product corresponding to a first resolution as the ground truth and the optimal remote sensing vegetation cover inversion product corresponding to a second resolution (excluding the first resolution) as input data, where the first resolution is higher than the second resolution; and training a neural network using the training dataset to obtain a target neural network corresponding to the second resolution. This invention can significantly improve the spatiotemporal resolution and accuracy of remote sensing vegetation cover inversion products.

Description

Technical Field

[0001] This invention relates to the field of remote sensing data fusion and vegetation cover monitoring technology, and in particular to a method, apparatus and equipment for fusion of multi-source remote sensing vegetation cover inversion products. Background Technology

[0002] Currently, vegetation cover monitoring methods are broadly categorized into two types: ground-based measurements and remote sensing. Ground-based measurements are valuable for small-scale, high-precision vegetation cover studies; however, due to environmental conditions, time constraints, and cost limitations, their spatial coverage is limited, making it difficult to quickly cover large areas. Therefore, for large-scale studies, satellite remote sensing monitoring has become an irreplaceable and crucial tool. Medium-resolution satellite data, represented by the Fengyun-3 series and the MODIS (Moderate-resolution Imaging Spectroradiometer) series, offers high update frequency and wide observation range, making it suitable for monitoring large-scale, rapidly changing vegetation dynamics. However, its relatively low spatial resolution (250m-1km) limits its application in small-scale detail analysis. High-resolution satellite data, represented by the Gaofen series, Landsat series, and Sentinel series, offers high spatial resolution (2m-30m), suitable for refined vegetation monitoring. However, its long repetition cycle limits its ability to monitor rapidly changing vegetation dynamics. Therefore, integrating the spatiotemporal advantages of multi-source satellites to obtain high spatiotemporal resolution vegetation cover monitoring results has become a key means to solve this problem.

[0003] In recent years, significant progress has been made in remote sensing vegetation cover fusion technology, mainly including spatiotemporal fusion models, machine learning models, and deep learning models. Spatiotemporal fusion models, such as STARFM (Starfield Factor Method), ESTARFM (Enhanced STARFM Algorithm), and FSDAF (Flexible Spatiotemporal Data Fusion), utilize MODIS and Landsat data to generate high-resolution images through spatial and temporal weighting, but they still have shortcomings in processing frequently changing vegetation cover data. Machine learning models, such as Random Forest and Support Vector Machine, have achieved some success in multi-source remote sensing data fusion, but they face performance bottlenecks when handling high-dimensional and complex scenes. Deep learning models, such as Convolutional Neural Networks (CNN) and Generative Adversarial Networks (GAN), have performed well in remote sensing image processing and data fusion, but their high training complexity and instability make them difficult to meet the needs of practical applications.

[0004] Most of the above studies are image-level fusions, which monitor vegetation cover based on the fused satellite remote sensing image data. However, this method fundamentally changes the observed spectral data, which may lead to inaccurate vegetation cover monitoring results. Summary of the Invention

[0005] In view of this, the purpose of the present invention is to provide a method, apparatus and equipment for fusing multi-source remote sensing vegetation cover inversion products, which can significantly improve the spatiotemporal resolution and accuracy of remote sensing vegetation cover inversion products.

[0006] In a first aspect, embodiments of the present invention provide a method for fusing multi-source remote sensing vegetation cover inversion products, including:

[0007] Obtain multi-source remote sensing vegetation cover inversion products, which include multiple remote sensing vegetation cover inversion products corresponding to different resolutions;

[0008] The authenticity of the remote sensing vegetation cover inversion products is verified so as to select the optimal remote sensing vegetation cover inversion product for each resolution from multiple remote sensing vegetation cover inversion products corresponding to each resolution.

[0009] The optimal remote sensing vegetation cover inversion product corresponding to the first resolution is used as the ground truth, and the optimal remote sensing vegetation cover inversion product corresponding to the second resolution (excluding the first resolution) is used as the input data to construct a training dataset, where the first resolution is higher than the second resolution.

[0010] The neural network is trained using the training dataset to obtain the target neural network corresponding to the second resolution;

[0011] The target neural network is used to generate an optimized result of the inversion product corresponding to the first resolution based on the remote sensing vegetation cover inversion product to be processed corresponding to the second resolution. The optimized result of the inversion product is used to fuse with the remote sensing vegetation cover inversion product to be processed corresponding to the first resolution to obtain a vegetation cover fusion product.

[0012] In one implementation, the authenticity of the remote sensing vegetation cover inversion product is verified to select the optimal remote sensing vegetation cover inversion product for each resolution from multiple remote sensing vegetation cover inversion products corresponding to each resolution, including:

[0013] Obtain the latitude and longitude information and measured vegetation coverage values ​​corresponding to the ground stations;

[0014] Using latitude and longitude information, the remote sensing vegetation cover inversion product and the measured vegetation cover value are spatiotemporally matched. The measured vegetation cover value obtained by spatiotemporal matching is used to verify the authenticity of the remote sensing vegetation cover inversion product and obtain the comprehensive quality evaluation score corresponding to the remote sensing vegetation cover inversion product.

[0015] For each resolution, based on the comprehensive quality evaluation score, the optimal remote sensing vegetation cover inversion product corresponding to that resolution is selected from multiple remote sensing vegetation cover inversion products.

[0016] In one implementation, the authenticity of the remote sensing vegetation cover inversion product is verified using spatiotemporally matched measured vegetation cover values, resulting in a comprehensive quality evaluation score for the remote sensing vegetation cover inversion product, including:

[0017] Based on the measured vegetation cover values ​​matched in time and space, multiple evaluation indicators corresponding to the remote sensing vegetation cover inversion products are determined.

[0018] The evaluation indicators are standardized to obtain multiple standardized evaluation indicators corresponding to the remote sensing vegetation cover inversion product.

[0019] The weight value corresponding to each standardized evaluation indicator is determined, and the standardized evaluation indicators are integrated using the weight value to obtain the comprehensive quality evaluation score corresponding to the remote sensing vegetation coverage inversion product.

[0020] In one implementation, the evaluation indicators include root mean square error (RMSE), correlation coefficient, and deviation, with the weights of the RMSE and deviation being higher than the weight of the correlation coefficient.

[0021] In one implementation, the overall quality assessment score is negatively correlated with the quality of the remote sensing vegetation cover inversion product; based on the overall quality assessment score, the optimal remote sensing vegetation cover inversion product corresponding to that resolution is selected from multiple remote sensing vegetation cover inversion products at that resolution, including:

[0022] Based on the comprehensive quality evaluation scores from low to high, the optimal remote sensing vegetation cover inversion product corresponding to this resolution is selected from multiple remote sensing vegetation cover inversion products corresponding to this resolution.

[0023] In one implementation, the neural network employs a U-Net network; the neural network is trained using a training dataset to obtain the target neural network corresponding to the second resolution, including:

[0024] Using the U-Net network, based on the optimal remote sensing vegetation cover inversion product corresponding to the second resolution, climate element data, topographic element data, and social element data, the optimized result of the inversion product corresponding to the first resolution is generated.

[0025] The loss value is determined based on the optimal remote sensing vegetation cover inversion product corresponding to the first resolution and the optimization result of the inversion product. The network parameters of the U-Net network are adjusted using the loss value to obtain the target neural network corresponding to the second resolution.

[0026] In one implementation, using a U-Net network, based on the optimal remote sensing vegetation cover inversion product corresponding to the second resolution, climate element data, topographic element data, and social element data, an optimized result of the inversion product corresponding to the first resolution is generated, including:

[0027] The encoder in the U-Net network performs convolution and pooling operations on the optimal remote sensing vegetation cover inversion product, climate element data, topographic element data and social element data corresponding to the second resolution to obtain a vegetation cover feature map.

[0028] By using the decoder in the U-Net network to perform deconvolution and skip connection operations on the vegetation cover feature map, the optimized result of the inversion product corresponding to the first resolution is obtained.

[0029] Secondly, embodiments of the present invention also provide a multi-source remote sensing vegetation cover inversion product fusion device, comprising:

[0030] The product acquisition module is used to acquire multi-source remote sensing vegetation cover inversion products, which include multiple remote sensing vegetation cover inversion products corresponding to different resolutions.

[0031] The product screening module is used to verify the authenticity of remote sensing vegetation cover inversion products, so as to select the optimal remote sensing vegetation cover inversion product for each resolution from multiple remote sensing vegetation cover inversion products corresponding to each resolution.

[0032] The training set construction module is used to construct a training dataset by taking the optimal remote sensing vegetation cover inversion product corresponding to the first resolution as the ground value and the optimal remote sensing vegetation cover inversion product corresponding to the second resolution other than the first resolution as the input data. The first resolution is higher than the second resolution.

[0033] The model training module is used to train the neural network using the training dataset to obtain the target neural network corresponding to the second resolution;

[0034] The target neural network is used to generate an optimized result of the inversion product corresponding to the first resolution based on the remote sensing vegetation cover inversion product to be processed corresponding to the second resolution. The optimized result of the inversion product is used to fuse with the remote sensing vegetation cover inversion product to be processed corresponding to the first resolution to obtain a vegetation cover fusion product.

[0035] Thirdly, embodiments of the present invention also provide an electronic device, including a processor and a memory, wherein the memory stores computer-executable instructions that can be executed by the processor, and the processor executes the computer-executable instructions to implement any of the methods provided in the first aspect.

[0036] Fourthly, embodiments of the present invention also provide a computer-readable storage medium storing computer-executable instructions, which, when invoked and executed by a processor, cause the processor to implement any of the methods provided in the first aspect.

[0037] The multi-source remote sensing vegetation cover inversion product fusion method, apparatus, and device provided in this invention first acquire multi-source remote sensing vegetation cover inversion products, which include multiple remote sensing vegetation cover inversion products corresponding to different resolutions. Then, the authenticity of the remote sensing vegetation cover inversion products is verified to select the optimal remote sensing vegetation cover inversion product for each resolution from the multiple remote sensing vegetation cover inversion products corresponding to each resolution. Next, a training dataset is constructed using the optimal remote sensing vegetation cover inversion product corresponding to the first resolution as the ground truth and the optimal remote sensing vegetation cover inversion product corresponding to the second resolution (excluding the first resolution) as input data, where the first resolution is higher than the second resolution. Finally, a neural network is trained using the training dataset to obtain a target neural network corresponding to the second resolution. The target neural network is used to generate an optimized result of the inversion product corresponding to the first resolution based on the remote sensing vegetation cover inversion product to be processed corresponding to the second resolution. The optimized result of the inversion product is then fused with the remote sensing vegetation cover inversion product to be processed corresponding to the first resolution to obtain a fused vegetation cover product. The above method selects the optimal remote sensing vegetation cover inversion product from remote sensing vegetation cover inversion products of different resolutions through authenticity verification, and uses it to construct a training dataset. This allows the trained target neural network to optimize the low-resolution remote sensing vegetation cover inversion products to be processed and fuse them with the high-resolution remote sensing vegetation cover inversion products to be processed. Compared with existing technologies that fundamentally change the observed spectral data, the embodiments of this invention realize remote sensing vegetation cover monitoring at the product level, significantly improving the spatiotemporal resolution and accuracy of remote sensing vegetation cover inversion products.

[0038] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention are realized and obtained in accordance with the structures particularly pointed out in the description, claims and drawings.

[0039] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0040] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0041] Figure 1 This is a flowchart illustrating a method for fusing multi-source remote sensing vegetation cover inversion products according to an embodiment of the present invention.

[0042] Figure 2 This is a schematic diagram of the overall process of a multi-source remote sensing vegetation cover inversion product fusion method provided in an embodiment of the present invention;

[0043] Figure 3 This is a schematic diagram of the structure of a multi-source remote sensing vegetation cover inversion product fusion device provided in an embodiment of the present invention;

[0044] Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0045] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0046] Currently, existing technologies have fundamentally altered the observed spectral data, which may lead to inaccuracies in vegetation cover monitoring results. Based on this, the present invention provides a method, apparatus, and equipment for fusing multi-source remote sensing vegetation cover inversion products, which can significantly improve the spatiotemporal resolution and accuracy of remote sensing vegetation cover inversion products.

[0047] To facilitate understanding of this embodiment, a detailed description of the multi-source remote sensing vegetation cover inversion product fusion method disclosed in this embodiment of the invention will be provided first. (See [link to relevant documentation]). Figure 1 The diagram shows a process flow chart for fusing multi-source remote sensing vegetation cover inversion products. This method mainly includes the following steps S102 to S108:

[0048] Step S102: Obtain multi-source remote sensing vegetation cover inversion products. The multi-source remote sensing vegetation cover inversion products include multiple remote sensing vegetation cover inversion products corresponding to different resolutions.

[0049] For example, it includes remote sensing vegetation cover inversion products corresponding to at least two resolutions, such as medium resolution and high resolution. Remote sensing vegetation cover inversion products corresponding to medium resolution include FY-3D, FY-3F, MODIS and other products, while remote sensing vegetation cover inversion products corresponding to high resolution include Landsat8 / 9, Sentinel-2A / B, GF-1 / 6 and other products.

[0050] Step S104: Verify the authenticity of the remote sensing vegetation cover inversion product to select the optimal remote sensing vegetation cover inversion product for each resolution from multiple remote sensing vegetation cover inversion products corresponding to each resolution.

[0051] In one example, by verifying the authenticity of the remote sensing vegetation cover inversion product, a comprehensive quality evaluation score can be obtained. The comprehensive quality evaluation score is negatively correlated with the quality of the remote sensing vegetation cover inversion product. Therefore, for each resolution, the remote sensing vegetation cover inversion product with the lowest comprehensive quality evaluation score at that resolution can be considered the optimal remote sensing vegetation cover inversion product. For example, the optimal remote sensing vegetation cover inversion product for medium resolution and the optimal remote sensing vegetation cover inversion product for high resolution can be selected separately.

[0052] Step S106: Using the optimal remote sensing vegetation cover inversion product corresponding to the first resolution as the ground truth, and the optimal remote sensing vegetation cover inversion product corresponding to the second resolution (other than the first resolution) as the input data, construct a training dataset.

[0053] The first resolution is higher than the second resolution. For example, the first resolution can be the aforementioned high resolution, and the second resolution can be the aforementioned medium resolution. The input data of the training dataset includes the optimal remote sensing vegetation cover inversion product corresponding to the medium resolution, climate element data, topographic element data, and social element data. The ground truth is the optimal remote sensing vegetation cover inversion product corresponding to the high resolution.

[0054] Step S108: Train the neural network using the training dataset to obtain the target neural network corresponding to the second resolution.

[0055] The target neural network can be a U-Net network. In one example, the aforementioned input data is fed into the U-Net network to obtain the optimized inversion product generated by the U-Net network based on the input data. The optimized inversion product result is then used to generate a loss function with the optimal remote sensing vegetation cover inversion product corresponding to the high resolution, in order to adjust the parameters of the U-Net network and obtain the trained target neural network. The target neural network is used to generate the optimized inversion product corresponding to the first resolution based on the remote sensing vegetation cover inversion product to be processed corresponding to the second resolution. The optimized inversion product result is then fused with the remote sensing vegetation cover inversion product to be processed corresponding to the first resolution to obtain a fused vegetation cover product.

[0056] The multi-source remote sensing vegetation cover inversion product fusion method provided in this invention verifies the authenticity of the data and selects the optimal remote sensing vegetation cover inversion product from remote sensing vegetation cover inversion products of different resolutions. This optimal product is then used to construct a training dataset. The trained target neural network optimizes the lower-resolution remote sensing vegetation cover inversion products to be processed and fuses them with the higher-resolution remote sensing vegetation cover inversion products to be processed. Compared with existing technologies that fundamentally change the observed spectral data, this invention achieves remote sensing vegetation cover monitoring at the product level, significantly improving the spatiotemporal resolution and accuracy of remote sensing vegetation cover inversion products.

[0057] To facilitate understanding, this invention provides a specific implementation method for fusing multi-source remote sensing vegetation cover inversion products. This invention aims to obtain high temporal resolution, high spatial resolution, and high precision vegetation cover fusion products by fusing vegetation cover products from multiple domestic and international sources with different spatiotemporal resolutions, such as FY-3F, MODIS, Landsat 8 / 9, Sentinel-2, and Gaofen-1 / 6 satellites, thereby providing strong data support for the refined monitoring of regional vegetation.

[0058] Specifically, the overall approach of this invention is as follows: collect publicly available measured vegetation cover values; collect vegetation cover inversion products from FY-3D, FY-3F, MODIS, Landsat 8 / 9, Sentinel-2A / B, and GF-1 / 6; normalize the multi-source vegetation cover inversion products; verify the authenticity of the multi-source satellite vegetation cover inversion products using measured vegetation cover values ​​as the true values; select a precision from medium-resolution satellites (FY-3D, FY-3F, MODIS) and high-resolution satellites (Landsat 8 / 9, Sentinel-2A / B, GF-1 / 6). The highest-resolution satellite vegetation cover inversion product is used in subsequent fusion. Using the best high-resolution satellite vegetation cover inversion product as the standard, and supplemented by climate, topography and social factors related to vegetation growth, a U-Net network model is established between the best high-resolution satellite vegetation cover inversion product (i.e. the best vegetation cover inversion product corresponding to the first resolution mentioned above) and the best medium-resolution satellite vegetation cover inversion product (i.e. the best vegetation cover inversion product corresponding to the second resolution mentioned above). This model is then applied to the best medium-resolution satellite vegetation cover inversion product to obtain a daily high-resolution and high-precision vegetation cover fusion product.

[0059] Based on this, this embodiment of the invention provides a specific implementation of the aforementioned step S104. The specific process for verifying the authenticity of multi-source satellite vegetation cover inversion data is shown in steps 1.1 to 1.3 below:

[0060] Step 1.1: Obtain the latitude and longitude information and measured vegetation coverage values ​​corresponding to the ground stations.

[0061] In one case, ground-measured vegetation cover data were downloaded from publicly available websites such as the National Ecological Data Center Resource Sharing Service Platform and the National Glacier, Permafrost and Desert Scientific Data Center. The data was analyzed, and quality control processes such as outlier removal were carried out to obtain the latitude and longitude information of the ground station and the measured vegetation cover values.

[0062] Furthermore, before performing spatiotemporal matching, the acquired remote sensing vegetation cover inversion products need to be preprocessed. The preprocessing process includes: collecting multi-source satellite vegetation cover inversion products, including vegetation cover data from medium-resolution satellites (FY-3D, FY-3F, MODIS) and high-resolution satellites (Landsat 8 / 9, Sentinel-2A / B, GF-1 / 6). The data is then analyzed to obtain parameters such as the observation range, resolution, and observation time of the satellite-inverted vegetation cover data. The vegetation cover inversion products generated by different satellites are normalized to ensure that data from different sources have the same numerical range and scale, and all multi-source satellite products are unified to an equal latitude and longitude projection.

[0063] Step 1.2: Use latitude and longitude information to perform spatiotemporal matching between the remote sensing vegetation cover inversion product and the measured vegetation cover value, so as to use the spatiotemporally matched measured vegetation cover value to verify the authenticity of the remote sensing vegetation cover inversion product and obtain the comprehensive quality evaluation score corresponding to the remote sensing vegetation cover inversion product.

[0064] In one example, the spatiotemporal matching process is as follows: using the measured vegetation cover value as the true value, based on the spatiotemporal information of the ground station and multi-source satellites, calculate the row and column number of the central latitude and longitude grid closest to the ground station, extract the vegetation cover inversion product within a 3*3 range based on the grid row and column number, take the mean as the satellite estimation result, and match it with the measured vegetation cover value.

[0065] In one example, after spatiotemporally matching the measured vegetation cover values ​​with satellite estimates, evaluation indicators are calculated to assess the quality of the vegetation cover product. The verification process is as follows:

[0066] (1) Based on the measured vegetation cover values ​​obtained through spatiotemporal matching, several evaluation indicators corresponding to the remote sensing vegetation cover inversion products are determined. Among them, the evaluation indicators include the root mean square error index, the correlation coefficient index, and the bias index.

[0067] 1) The deviation calculation formula is as follows:

[0068] Bias = x i -x oi ;

[0069] In the formula, Bias represents the average deviation; Xi represents the data to be tested, i.e., the satellite estimation result; and Xoi represents the source data for testing, i.e., the measured value of vegetation cover. The deviation is calculated for each matched sample pixel, and the obtained deviation value can be used to draw a deviation spatial distribution map.

[0070] 2) The formula for calculating the root mean square error is as follows:

[0071]

[0072] In the formula: RMSE represents the root mean square error; N represents the number of matched samples; Xi represents the data to be tested; Xoi represents the source data for testing.

[0073] 3) The formula for calculating the correlation coefficient is shown below:

[0074]

[0075] In the formula, Corr represents the correlation coefficient; N represents the number of matched samples; Xi represents the data to be tested; and Xoi represents the source data for the test. This represents the sample mean of the data to be tested; This represents the sample mean of the test data.

[0076] (2) The evaluation indicators are standardized to obtain multiple standardized evaluation indicators corresponding to the remote sensing vegetation coverage inversion product.

[0077] To comprehensively evaluate the quality of multi-source satellite vegetation cover products, the following comprehensive evaluation method is adopted, which integrates and analyzes three indicators: root mean square error (RMSE), correlation coefficient (R), and bias.

[0078] The root mean square error (RMSE), correlation coefficient (R), and bias are standardized to ensure consistent dimensions. The following formula is used to standardize the indicators:

[0079] For RMSE and Bias, the smaller the standardized values, the better:

[0080]

[0081] For the correlation coefficient R, a larger standardized value is better:

[0082]

[0083] Where X is the original value, X norn X is the standardized value. min X max These are the minimum and maximum values, respectively.

[0084] (3) Determine the weight value corresponding to each standardized evaluation indicator, so as to use the weight value to fuse the standardized evaluation indicators and obtain the comprehensive quality evaluation score corresponding to the remote sensing vegetation coverage inversion product.

[0085] In one example, the weights of the root mean square error (RMSE) and bias indicators are higher than the weight of the correlation coefficient. Generally, RMSE and bias have higher weights because they directly reflect product error; while the correlation coefficient reflects the correlation between the product and ground measurements, and its weight can be slightly lower. In this embodiment of the invention, the weights of RMSE and bias are both 0.4, and the weight of the correlation coefficient is 0.2.

[0086] In one example, a weighted summation method is used to comprehensively calculate the standardized indicators:

[0087] Score = 0.4 × RMSE norm +0.4×Bias norm +0.2×R norm ;

[0088] Among them, "Score" refers to the overall quality evaluation score.

[0089] Step 1.3: For each resolution, based on the comprehensive quality evaluation score, select the optimal remote sensing vegetation cover inversion product corresponding to that resolution from among multiple remote sensing vegetation cover inversion products corresponding to that resolution.

[0090] In one example, the overall quality assessment score is negatively correlated with the quality of the remote sensing vegetation cover inversion product; that is, the lower the overall quality assessment score, the better the overall quality of the product. Therefore, the optimal remote sensing vegetation cover inversion product for a given resolution can be selected from multiple remote sensing vegetation cover inversion products corresponding to that resolution, arranged in ascending order of overall quality assessment score.

[0091] Specifically, based on the comprehensive quality evaluation score, the vegetation cover inversion product with the lowest comprehensive quality evaluation score is selected from the medium-resolution satellite products as the optimal medium-resolution vegetation cover inversion product; similarly, the vegetation cover inversion product with the lowest comprehensive quality evaluation score is selected from the high-resolution satellite products as the optimal high-resolution vegetation cover inversion product. These two optimal products will participate in the subsequent fusion calculation.

[0092] Furthermore, the neural network in this embodiment of the invention employs the U-Net network. The U-Net network model is chosen for fusing multi-source remote sensing vegetation cover data primarily due to its powerful multi-scale feature extraction and fusion capabilities. U-Net is a fully convolutional network (FCN), mainly composed of an encoder (Contracting Path) and a decoder (Expanding Path). It effectively combines high-resolution and low-resolution features through skip connections, enhancing the preservation of detailed information. In addition, U-Net supports end-to-end training, reducing intermediate processing steps and improving processing efficiency; it also performs excellently in few-shot learning and data augmentation, achieving good results with limited training data and addressing the problem of scarce high-quality labeled data.

[0093] Based on this, this invention provides a specific implementation method for constructing a training dataset, including: preprocessing the optimal medium-resolution satellite vegetation cover inversion product, the optimal high-resolution satellite vegetation cover inversion product, and related climate element data, topographic element data, and social element data; adjusting their value range to between [0,1] through normalization processing to generate a training dataset suitable for the U-Net network. The optimal high-resolution satellite vegetation cover inversion product is used as the label data (i.e., the ground truth).

[0094] This embodiment of the invention further provides a specific implementation method for training a U-Net network, including the following steps 2.1 to 2.2:

[0095] Step 2.1: Using the U-Net network, based on the optimal remote sensing vegetation cover inversion product corresponding to the second resolution, climate element data, topographic element data, and social element data, generate the optimized result of the inversion product corresponding to the first resolution. Specifically:

[0096] (1) Through the encoder in the U-Net network, convolution and pooling operations are performed on the optimal remote sensing vegetation cover inversion product, climate element data, topographic element data and social element data corresponding to the second resolution to obtain the vegetation cover feature map.

[0097] The encoder part of the U-Net network: The encoder extracts high-level features of the input data through a series of convolutional layers and max pooling layers.

[0098] The expression for the convolution operation is as follows:

[0099] Conv(X) = W*X + b;

[0100] Where W is the kernel weight, b is the bias, X is the input feature map, and * is the convolution operation.

[0101] The expression for pooling operations (max pooling) is shown below:

[0102]

[0103] Here, window is the pooled window, and max represents the maximum value within the window.

[0104] (2) By using the decoder in the U-Net network, the vegetation coverage feature map is deconvolutionally processed and skip connection is performed to obtain the optimization result of the inversion product corresponding to the first resolution.

[0105] The decoder part of the U-Net network: The decoder gradually restores the resolution of the image through a series of deconvolution (upsampling) operations, and finally outputs a vegetation cover prediction map with the same resolution as the label data.

[0106] The expression for the deconvolution operation is shown below:

[0107] DeConv(X) = WT*X + b;

[0108] Among them, W T This is a transposed convolution kernel.

[0109] Skip connections: One of the key features of U-Net, skip connections directly pass the feature maps from the encoder to the corresponding layers of the decoder, preserving more detailed information. The expression for a skip connection is shown below:

[0110] SkipConnection(X encoder X decoder =Concat(X) encoder X decoder ).

[0111] Step 2.2: Determine the loss value based on the optimal remote sensing vegetation cover inversion product corresponding to the first resolution and the optimization result of the inversion product, so as to adjust the network parameters of the U-Net network using the loss value and obtain the target neural network corresponding to the second resolution.

[0112] In one example, a loss function is defined to evaluate the difference between the model output and the actual labels. Commonly used loss functions include mean squared error loss (MSE) and cross-entropy loss.

[0113] For example, the mean squared error loss function is shown below:

[0114]

[0115] Among them, Y i If it is a true label value, is the model's predicted value, and N is the number of samples.

[0116] In one example, during the training phase, the model parameters are continuously adjusted using the backpropagation algorithm and an optimization algorithm (such as the Adam optimizer). The expression for backpropagation is shown below:

[0117]

[0118] Where θ is the model parameter, η is the learning rate, and L(θ) is the loss function. For gradient.

[0119] In one example, after training is complete, the model is evaluated using a test dataset, and metrics such as root mean square error (RMSE) and correlation coefficient (R) are calculated to verify the model's performance.

[0120]

[0121] Where X represents the input data. This represents the fusion result of the predicted vegetation cover.

[0122] Furthermore, after the U-Net network is trained, it can be used to convert the medium-resolution satellite vegetation cover inversion product to be processed into a high-resolution optimized product. This optimized product is then fused with the acquired high-resolution satellite vegetation cover inversion product to obtain the final fused vegetation cover product. In one example, applying this model to daily data can generate daily high-resolution, high-precision fused vegetation cover products.

[0123] In summary, the multi-source remote sensing vegetation cover inversion product fusion method provided in this embodiment of the invention has at least the following characteristics:

[0124] (1) Improving the spatial resolution and accuracy of vegetation cover products: By utilizing the U-Net network model, the optimal medium-resolution satellite vegetation cover products are fused with the optimal high-resolution satellite products, effectively improving the spatial resolution and accuracy of the final vegetation cover products. This allows the generated vegetation cover products to reflect the actual condition of the surface vegetation in greater detail, making them suitable for ecological environment monitoring and research with higher precision requirements.

[0125] (2) Improved spatiotemporal consistency of data: The fusion method used in this invention not only improves spatial resolution but also maintains data consistency over time, generating daily high-resolution vegetation cover products. Compared to traditional single data sources or low-frequency products, the results of this invention can provide more frequent and accurate time-series data for monitoring vegetation dynamic changes.

[0126] (3) Enhancing the efficiency of comprehensive utilization of multi-source data: By integrating vegetation cover products from different satellite platforms, this invention maximizes the advantages of various types of satellite data. Medium-resolution satellites provide a relatively wide coverage area, while high-resolution satellites provide fine spatial details. The combination of the two provides users with comprehensive vegetation cover data that combines macroscopic and microscopic perspectives.

[0127] (4) Adapting to the diverse needs of different ecological environments: This invention considers various factors related to vegetation growth, such as climate, topography, and social factors, and constructs a highly adaptable fusion model. This model can be adjusted according to the actual needs of different ecological environments to generate more targeted vegetation cover products, providing strong data support for the management and protection of different ecosystems.

[0128] (5) Enhancing the model's versatility and scalability: The application of the U-Net network model enables this invention to have good versatility and scalability in applications across different regions and scales. By continuously updating training data and adjusting model parameters, this method can adapt to environmental conditions in different regions, ensuring that the generated vegetation cover products have high reliability in various application scenarios.

[0129] Based on the foregoing embodiments, the embodiments of the present invention further provide, as follows: Figure 2 The diagram illustrates the overall process of a multi-source remote sensing vegetation cover inversion product fusion method, including: using measured vegetation cover values ​​to perform authenticity checks on medium-resolution remote sensing vegetation cover inversion products (FY-3D, FY-3F, MODIS) and high-resolution remote sensing vegetation cover inversion products (Landsat8 / 9, Sentinel-2A / B, GF-1 / 6), obtaining a comprehensive quality evaluation score, and then selecting the optimal medium-resolution satellite vegetation cover inversion product and the optimal high-resolution satellite vegetation cover inversion product; combining land use classification, GDP, population, temperature, precipitation, altitude, slope, and other data for spatiotemporal matching; inputting the spatiotemporally matched data into the U-Net network to calculate high spatiotemporal resolution vegetation cover, and outputting a high spatiotemporal resolution vegetation cover product.

[0130] In summary, the purpose of this invention is to address the limitations of spatial and temporal resolution in current vegetation cover monitoring. By proposing a multi-source remote sensing vegetation cover product fusion technology based on the U-Net network, it aims to leverage the spatiotemporal advantages of data from Fengyun-3, MODIS, Landsat, Gaofen, and Sentinel satellites to generate high spatiotemporal resolution vegetation cover data. The U-Net network possesses efficient image segmentation capabilities and the ability to handle complex scenes. Through deep learning fusion of different satellite data, it maintains the accuracy of the original product data while improving the resolution of the monitoring results. This meets the needs of large-scale and rapidly changing vegetation dynamic monitoring, providing more accurate vegetation cover assessments and strong data support for refined regional vegetation monitoring.

[0131] Based on the foregoing embodiments, this invention provides a multi-source remote sensing vegetation cover inversion product fusion device, see [link to relevant documentation]. Figure 3 The diagram shows a structural schematic of a multi-source remote sensing vegetation cover inversion product fusion device, which mainly includes the following parts:

[0132] Product acquisition module 302 is used to acquire multi-source remote sensing vegetation cover inversion products, which include multiple remote sensing vegetation cover inversion products corresponding to different resolutions.

[0133] Product screening module 304 is used to verify the authenticity of remote sensing vegetation cover inversion products, so as to select the optimal remote sensing vegetation cover inversion product corresponding to each resolution from multiple remote sensing vegetation cover inversion products corresponding to each resolution.

[0134] Training set construction module 306 is used to construct a training dataset by taking the optimal remote sensing vegetation cover inversion product corresponding to the first resolution as the ground value and the optimal remote sensing vegetation cover inversion product corresponding to the second resolution other than the first resolution as the input data. The first resolution is higher than the second resolution.

[0135] The model training module 308 is used to train the neural network using the training dataset to obtain the target neural network corresponding to the second resolution.

[0136] The target neural network is used to generate an optimized result of the inversion product corresponding to the first resolution based on the remote sensing vegetation cover inversion product to be processed corresponding to the second resolution. The optimized result of the inversion product is used to fuse with the remote sensing vegetation cover inversion product to be processed corresponding to the first resolution to obtain a vegetation cover fusion product.

[0137] The multi-source remote sensing vegetation cover inversion product fusion device provided in this invention verifies the authenticity of remote sensing vegetation cover inversion products of different resolutions to select the optimal remote sensing vegetation cover inversion product, which is then used to construct a training dataset. This dataset enables the trained target neural network to optimize the lower-resolution remote sensing vegetation cover inversion product to be processed and fuse it with the higher-resolution remote sensing vegetation cover inversion product to be processed. Compared with existing technologies that fundamentally change the observed spectral data, this invention realizes remote sensing vegetation cover monitoring at the product level, significantly improving the spatiotemporal resolution and accuracy of remote sensing vegetation cover inversion products.

[0138] In one implementation, the product screening module 304 is specifically used for:

[0139] Obtain the latitude and longitude information and measured vegetation coverage values ​​corresponding to the ground stations;

[0140] Using latitude and longitude information, the remote sensing vegetation cover inversion product and the measured vegetation cover value are spatiotemporally matched. The measured vegetation cover value obtained by spatiotemporal matching is used to verify the authenticity of the remote sensing vegetation cover inversion product and obtain the comprehensive quality evaluation score corresponding to the remote sensing vegetation cover inversion product.

[0141] For each resolution, based on the comprehensive quality evaluation score, the optimal remote sensing vegetation cover inversion product corresponding to that resolution is selected from multiple remote sensing vegetation cover inversion products.

[0142] In one implementation, the product screening module 304 is specifically used for:

[0143] Based on the measured vegetation cover values ​​matched in time and space, multiple evaluation indicators corresponding to the remote sensing vegetation cover inversion products are determined.

[0144] The evaluation indicators are standardized to obtain multiple standardized evaluation indicators corresponding to the remote sensing vegetation cover inversion product.

[0145] The weight value corresponding to each standardized evaluation indicator is determined, and the standardized evaluation indicators are integrated using the weight value to obtain the comprehensive quality evaluation score corresponding to the remote sensing vegetation coverage inversion product.

[0146] In one implementation, the evaluation indicators include root mean square error (RMSE), correlation coefficient, and deviation, with the weights of the RMSE and deviation being higher than the weight of the correlation coefficient.

[0147] In one implementation, the overall quality evaluation score is negatively correlated with the quality of the remote sensing vegetation cover inversion product; the product screening module 304 is specifically used for:

[0148] Based on the comprehensive quality evaluation scores from low to high, the optimal remote sensing vegetation cover inversion product corresponding to this resolution is selected from multiple remote sensing vegetation cover inversion products corresponding to this resolution.

[0149] In one implementation, the neural network employs a U-Net network; the model training module 308 is specifically used for:

[0150] Using the U-Net network, based on the optimal remote sensing vegetation cover inversion product corresponding to the second resolution, climate element data, topographic element data, and social element data, the optimized result of the inversion product corresponding to the first resolution is generated.

[0151] The loss value is determined based on the optimal remote sensing vegetation cover inversion product corresponding to the first resolution and the optimization result of the inversion product. The network parameters of the U-Net network are adjusted using the loss value to obtain the target neural network corresponding to the second resolution.

[0152] In one implementation, the model training module 308 is specifically used for:

[0153] The encoder in the U-Net network performs convolution and pooling operations on the optimal remote sensing vegetation cover inversion product, climate element data, topographic element data and social element data corresponding to the second resolution to obtain a vegetation cover feature map.

[0154] By using the decoder in the U-Net network to perform deconvolution and skip connection operations on the vegetation cover feature map, the optimized result of the inversion product corresponding to the first resolution is obtained.

[0155] The device provided in this embodiment of the invention has the same implementation principle and technical effect as the aforementioned method embodiment. For the sake of brevity, any parts not mentioned in the device embodiment can be referred to the corresponding content in the aforementioned method embodiment.

[0156] This invention provides an electronic device, specifically, the electronic device includes a processor and a storage device; the storage device stores a computer program, and the computer program, when run by the processor, executes the method described in any of the above embodiments.

[0157] Figure 4 The present invention provides a schematic diagram of the structure of an electronic device 100, which includes a processor 40, a memory 41, a bus 42 and a communication interface 43. The processor 40, the communication interface 43 and the memory 41 are connected through the bus 42. The processor 40 is used to execute executable modules, such as computer programs, stored in the memory 41.

[0158] The memory 41 may include high-speed random access memory (RAM) or non-volatile memory, such as at least one disk storage device. Communication between this system network element and at least one other network element is achieved through at least one communication interface 43 (which can be wired or wireless), such as the Internet, wide area network, local area network, metropolitan area network, etc.

[0159] Bus 42 can be an ISA bus, PCI bus, or EISA bus, etc. The bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 4 The symbol is represented by a single double-headed arrow, but this does not mean that there is only one bus or one type of bus.

[0160] The memory 41 is used to store programs. After receiving an execution instruction, the processor 40 executes the program. The method executed by the device for defining the flow process disclosed in any of the foregoing embodiments of the present invention can be applied to the processor 40 or implemented by the processor 40.

[0161] Processor 40 may be an integrated circuit chip with signal processing capabilities. In implementation, each step of the above method can be completed by the integrated logic circuitry in the hardware of processor 40 or by instructions in software form. Processor 40 can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it can also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this invention. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this invention can be directly embodied in the execution of a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor. The software modules can reside in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. The storage medium is located in memory 41. The processor 40 reads the information in memory 41 and, in conjunction with its hardware, completes the steps of the above method.

[0162] The computer program product of the readable storage medium provided in the embodiments of the present invention includes a computer-readable storage medium storing program code. The instructions included in the program code can be used to execute the methods described in the foregoing method embodiments. For specific implementation, please refer to the foregoing method embodiments, which will not be repeated here.

[0163] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, essentially, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0164] Finally, it should be noted that the above-described embodiments are merely specific implementations of the present invention, used to illustrate the technical solutions of the present invention, and not to limit it. The scope of protection of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments within the technical scope disclosed in the present invention, or make equivalent substitutions for some of the technical features; and these modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for fusing multi-source remote sensing vegetation cover inversion products, characterized in that, include: Obtain multi-source remote sensing vegetation cover inversion products, which include multiple remote sensing vegetation cover inversion products corresponding to different resolutions; The authenticity of the remote sensing vegetation cover inversion product is verified in order to select the optimal remote sensing vegetation cover inversion product corresponding to each resolution from the multiple remote sensing vegetation cover inversion products corresponding to each resolution. Using the optimal remote sensing vegetation cover inversion product corresponding to the first resolution as the ground truth, and the optimal remote sensing vegetation cover inversion product corresponding to the second resolution other than the first resolution as the input data, a training dataset is constructed, wherein the first resolution is higher than the second resolution. The neural network is trained using the training dataset to obtain the target neural network corresponding to the second resolution; The target neural network is used to generate an optimized result of the inversion product corresponding to the first resolution based on the remote sensing vegetation cover inversion product to be processed corresponding to the second resolution. The optimized result of the inversion product is used to fuse with the remote sensing vegetation cover inversion product to be processed corresponding to the first resolution to obtain a vegetation cover fusion product. The authenticity verification of the remote sensing vegetation cover inversion products is performed to select the optimal remote sensing vegetation cover inversion product corresponding to each resolution from multiple remote sensing vegetation cover inversion products corresponding to each resolution. This includes: acquiring the latitude and longitude information and measured vegetation cover values ​​corresponding to ground stations; performing spatiotemporal matching of the remote sensing vegetation cover inversion products and the measured vegetation cover values ​​using the latitude and longitude information, and using the spatiotemporally matched measured vegetation cover values ​​to verify the authenticity of the remote sensing vegetation cover inversion products, thereby obtaining a comprehensive quality evaluation score corresponding to the remote sensing vegetation cover inversion products; for each resolution, based on the comprehensive quality evaluation score, selecting the optimal remote sensing vegetation cover inversion product corresponding to that resolution from multiple remote sensing vegetation cover inversion products corresponding to that resolution. The authenticity of the remote sensing vegetation cover inversion product is verified by using the spatiotemporally matched measured vegetation cover values ​​to obtain a comprehensive quality evaluation score corresponding to the remote sensing vegetation cover inversion product. This includes: determining multiple evaluation indicators corresponding to the remote sensing vegetation cover inversion product based on the spatiotemporally matched measured vegetation cover values; standardizing the evaluation indicators to obtain multiple standardized evaluation indicators corresponding to the remote sensing vegetation cover inversion product; determining the weight value corresponding to each standardized evaluation indicator, and using the weight value to fuse the standardized evaluation indicators to obtain a comprehensive quality evaluation score corresponding to the remote sensing vegetation cover inversion product. The neural network employs a U-Net network. Training the neural network using the training dataset to obtain the target neural network corresponding to the second resolution includes: generating an optimization result for the inversion product corresponding to the first resolution based on the optimal remote sensing vegetation cover inversion product, climate element data, topographic element data, and social element data corresponding to the second resolution using the U-Net network; determining a loss value based on the optimal remote sensing vegetation cover inversion product and the optimization result of the inversion product corresponding to the first resolution, and adjusting the network parameters of the U-Net network using the loss value to obtain the target neural network corresponding to the second resolution.

2. The method for fusing multi-source remote sensing vegetation cover inversion products according to claim 1, characterized in that, The evaluation indicators include root mean square error, correlation coefficient, and deviation. The weight values ​​of the root mean square error and deviation are higher than the weight value of the correlation coefficient.

3. The method for fusing multi-source remote sensing vegetation cover inversion products according to claim 1, characterized in that, The overall quality evaluation score is negatively correlated with the quality of the remote sensing vegetation cover inversion product; based on the overall quality evaluation score, the optimal remote sensing vegetation cover inversion product corresponding to this resolution is selected from multiple remote sensing vegetation cover inversion products corresponding to this resolution, including: Based on the comprehensive quality evaluation scores from low to high, the optimal remote sensing vegetation cover inversion product corresponding to that resolution is selected from multiple remote sensing vegetation cover inversion products corresponding to that resolution.

4. The method for fusing multi-source remote sensing vegetation cover inversion products according to claim 1, characterized in that, Using the U-Net network, based on the optimal remote sensing vegetation cover inversion product corresponding to the second resolution, the climate element data, the terrain element data, and the social element data, an optimized result of the inversion product corresponding to the first resolution is generated, including: The encoder in the U-Net network performs convolution and pooling operations on the optimal remote sensing vegetation cover inversion product, climate element data, terrain element data, and social element data corresponding to the second resolution to obtain a vegetation cover feature map. The vegetation cover feature map is deconvolutionally processed and skip connection operations are performed on the decoder in the U-Net network to obtain the optimized inversion product result corresponding to the first resolution.

5. A device for fusing multi-source remote sensing vegetation cover inversion products, characterized in that, include: The product acquisition module is used to acquire multi-source remote sensing vegetation cover inversion products, which include multiple remote sensing vegetation cover inversion products corresponding to different resolutions. The product screening module is used to verify the authenticity of the remote sensing vegetation coverage inversion product, so as to select the optimal remote sensing vegetation coverage inversion product corresponding to each resolution from the multiple remote sensing vegetation coverage inversion products corresponding to each resolution. The training set construction module is used to construct a training dataset by taking the optimal remote sensing vegetation cover inversion product corresponding to the first resolution as the ground value and the optimal remote sensing vegetation cover inversion product corresponding to the second resolution other than the first resolution as the input data, wherein the first resolution is higher than the second resolution. The model training module is used to train the neural network using the training dataset to obtain the target neural network corresponding to the second resolution; The target neural network is used to generate an optimized result of the inversion product corresponding to the first resolution based on the remote sensing vegetation cover inversion product to be processed corresponding to the second resolution. The optimized result of the inversion product is used to fuse with the remote sensing vegetation cover inversion product to be processed corresponding to the first resolution to obtain a vegetation cover fusion product. The product screening module is specifically used for: acquiring the latitude and longitude information and measured vegetation coverage values ​​corresponding to ground stations; using the latitude and longitude information to perform spatiotemporal matching between the remote sensing vegetation coverage inversion product and the measured vegetation coverage values, so as to use the spatiotemporally matched measured vegetation coverage values ​​to verify the authenticity of the remote sensing vegetation coverage inversion product and obtain the comprehensive quality evaluation score corresponding to the remote sensing vegetation coverage inversion product; for each resolution, based on the comprehensive quality evaluation score, selecting the optimal remote sensing vegetation coverage inversion product corresponding to that resolution from multiple remote sensing vegetation coverage inversion products corresponding to that resolution. The product screening module is specifically used for: determining multiple evaluation indicators corresponding to the remote sensing vegetation coverage inversion product based on the spatiotemporally matched measured vegetation coverage values; standardizing the evaluation indicators to obtain multiple standardized evaluation indicators corresponding to the remote sensing vegetation coverage inversion product; determining the weight value corresponding to each standardized evaluation indicator, and using the weight value to fuse the standardized evaluation indicators to obtain the comprehensive quality evaluation score corresponding to the remote sensing vegetation coverage inversion product. The neural network adopts the U-Net network; the model training module is specifically used to: generate the optimization result of the inversion product corresponding to the first resolution based on the optimal remote sensing vegetation cover inversion product, climate element data, terrain element data and social element data corresponding to the second resolution through the U-Net network; Based on the optimal remote sensing vegetation cover inversion product corresponding to the first resolution and the optimization result of the inversion product, a loss value is determined, and the network parameters of the U-Net network are adjusted using the loss value to obtain the target neural network corresponding to the second resolution.

6. An electronic device, characterized in that, The method includes a processor and a memory, the memory storing computer-executable instructions executable by the processor, the processor executing the computer-executable instructions to implement the method of any one of claims 1 to 4.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions that, when invoked and executed by a processor, cause the processor to perform the method described in any one of claims 1 to 4.

Images