Remote sensing image surface water extraction method, device, system, and storage medium

Abstract

The present invention discloses a method, device, system, and storage medium for extracting surface water from remote sensing images, comprising: step S1, obtaining a remote sensing surface water dataset; step S2, constructing a remote sensing surface water extraction network (RSWE-Net) based on the remote sensing surface water dataset; step S3, dividing the remote sensing surface water dataset into a training set and a test set; step S4, training the remote sensing surface water extraction network (RSWE-Net) based on the training set; and step S5, inputting the test set into the trained remote sensing surface water extraction network (RSWE-Net) to extract surface water from remote sensing images. The technical solution of the present invention improves the accuracy of water extraction.

Description

Technical Field

[0001] The present invention belongs to the technical field of image processing, and in particular relates to a method and device, a system, and a storage medium for extracting surface water bodies from remote sensing images. Background Art

[0002] With the improvement of people's living standards, environmental protection has received increasing attention in recent years, and more and more people have begun to realize its importance. The spatial distribution and area of ​​surface water, as a key indicator of the state of the natural environment, not only reflects its characteristics but is also, to a certain extent, affected by global climate change, which is largely due to the increase in human activities. As a key component of the hydrological cycle, the accurate extraction of surface water bodies is of great significance in multiple fields, including water change detection, water resource protection, global climate change research, and flood disaster emergency response.

[0003] Traditional water segmentation methods primarily rely on thresholding water indices. However, when applied to high-resolution remote sensing imagery, these methods suffer from increased spectral variability due to the increased spatial detail within the images, leading to a decrease in their generalization ability. While machine learning methods can somewhat mitigate the errors introduced by threshold selection, most still rely solely on low-level features such as color, shape, and texture of surface water bodies to complete the segmentation process, often resulting in significant salt-and-pepper noise in the extracted results. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a method and device, system and storage medium for extracting surface water bodies from remote sensing images, so as to improve the accuracy of water body extraction.

[0005] To achieve the above object, the present invention adopts the following technical solutions:

[0006] A method for extracting surface water bodies from remote sensing images, comprising:

[0007] Step S1, obtaining a remote sensing surface water dataset;

[0008] Step S2: constructing a remote sensing surface water extraction network RSWE-Net based on the remote sensing surface water dataset; wherein the remote sensing surface water extraction network RSWE-Net includes: a cross-scale adaptive feature integration encoding module; a multi-level small target water body precision optimization module; and a contextual semantic attention decoding module;

[0009] Step S3, dividing the remote sensing surface water dataset to obtain a training set and a test set;

[0010] Step S4, training the remote sensing surface water extraction network RSWE-Net according to the training set;

[0011] Step S5: Input the test set into the trained remote sensing surface water extraction network RSWE-Net to extract surface water bodies in the remote sensing image.

[0012] Preferably, the cross-scale adaptive feature integration coding module includes: a hybrid multi-path pooling convolution sub-module, which extracts cross-scale adaptive feature maps in parallel with the pooling residual mechanism and ordinary convolution.

[0013] Preferably, the multi-level small target water body precision optimization module includes: a pooled convolution normalized activation submodule, which combines the horizontal and vertical extraction mechanisms and different types of hole convolutions in parallel to extract multi-level small target water body feature maps with different receptive fields.

[0014] Preferably, the contextual semantic attention decoding module learns and synthesizes different feature information through four different branches consisting of maximum pooling, average pooling, void convolution and ordinary convolution, and simultaneously splices the shallow feature map with the contextual semantic feature map in the channel dimension.

[0015] The present invention also provides a device for extracting surface water from a remote sensing image, comprising:

[0016] Acquisition module, used to obtain remote sensing surface water datasets;

[0017] A construction module is used to construct a remote sensing surface water extraction network RSWE-Net based on a remote sensing surface water dataset. The RSWE-Net includes: a cross-scale adaptive feature integration encoding module; a multi-level small target water accuracy optimization module; and a contextual semantic attention decoding module.

[0018] The partitioning module is used to partition the remote sensing surface water dataset into training and test sets;

[0019] The training module is used to train the remote sensing surface water extraction network RSWE-Net based on the training set;

[0020] The extraction module is used to input the test set into the trained remote sensing surface water extraction network RSWE-Net to extract surface water bodies in remote sensing images.

[0021] An embodiment of the present invention also provides a remote sensing image surface water body extraction system, comprising: a memory and a processor, wherein the memory stores a computer program run by the processor, and the computer program executes a remote sensing image surface water body extraction method when run by the processor.

[0022] An embodiment of the present invention further provides a storage medium having a computer program stored thereon. When the computer program is run, the method for extracting surface water bodies from remote sensing images is executed.

[0023] The present invention adopts a remote sensing surface water extraction network RSWE-Net, which includes: a cross-scale adaptive feature integration coding module, a multi-level small target water body precision optimization module, and a contextual semantic attention decoding module. The cross-scale adaptive feature integration coding module in RSWE-Net can dynamically adjust and fuse feature maps from different scales in an adaptive manner, which helps to capture detailed information of surface water bodies at different resolutions, such as small-area lakes and rivers and large-scale water body contours, thereby significantly enhancing the model's comprehensive representation ability of water body features; the multi-level small target water body precision optimization module can perform refined feature extraction for small target water bodies of different scales and complexities by constructing a multi-level network structure, enabling the module to capture subtle features of small water bodies such as streams and puddles, thereby significantly improving the detection accuracy of these small targets; the contextual semantic attention decoding module can dynamically focus on key semantic information in the input remote sensing image by introducing an attention mechanism, so that the model can more accurately understand the semantic relationship between water bodies and other land features during the decoding stage. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0025] Figure 1 This is a flow chart of a method for extracting surface water from remote sensing images according to an embodiment of the present invention;

[0026] Figure 2 Schematic diagram of RSWE-Net structure;

[0027] Figure 3 This is the data processing flow chart of RSWE-Net;

[0028] Figure 4 This is a schematic diagram of the HMPCB structure on branch A1;

[0029] Figure 5 This is a schematic diagram of the structure of the multi-level small target water body precision optimization module;

[0030] Figure 6 Schematic diagram of the PCNAB structure of the pooling convolution normalization activation submodule. DETAILED DESCRIPTION

[0031] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0032] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0033] Example 1:

[0034] like Figure 1 As shown, an embodiment of the present invention provides a method for extracting surface water bodies from remote sensing images, comprising:

[0035] Step S1, obtaining a remote sensing surface water dataset;

[0036] Step S2: constructing a remote sensing surface water extraction network RSWE-Net based on the remote sensing surface water dataset;

[0037] Step S3, dividing the remote sensing surface water dataset to obtain a training set and a test set;

[0038] Step S4, training the remote sensing surface water extraction network RSWE-Net according to the training set;

[0039] Step S5: Input the test set into the trained remote sensing surface water extraction network RSWE-Net to extract surface water bodies in the remote sensing image.

[0040] As an implementation method of an embodiment of the present invention, in step S1, first, each picture in the remote sensing image is adjusted to a uniform size to obtain a remote sensing surface water classification dataset; secondly, each picture in the remote sensing surface water classification dataset is labeled with surface water bodies to generate a remote sensing surface water segmentation dataset; then, the above-mentioned remote sensing surface water segmentation dataset is converted into a remote sensing surface water semantic segmentation dataset, and the RGB color label of each picture in the remote sensing surface water semantic segmentation dataset is converted into a corresponding remote sensing surface water single-channel digital label; finally, the single-channel digital label of the remote sensing surface water is accurately matched with the remote sensing surface water classification dataset, and each remote sensing image in the remote sensing surface water classification dataset is made to correspond to its unique, corresponding single-channel digital label, thereby forming a remote sensing surface water dataset.

[0041] In the experiment of the present invention, GF-6 remote sensing images of a certain area in Anhui Province were used. By using a Python script, the size of 3,000 images in the GF-6 remote sensing images was resized to 512×512 pixels to obtain a remote sensing surface water classification dataset. Labelme was used to manually annotate each image in the remote sensing surface water classification dataset to obtain a remote sensing surface water segmentation dataset, and the remote sensing surface water segmentation dataset was converted into a remote sensing surface water semantic segmentation dataset using the "labelme2voc" script, and the 3,000 RGB color labels in the semantic segmentation dataset were converted into corresponding remote sensing surface water single-channel digital labels. The final remote sensing surface water dataset consists of 3,000 remote sensing surface water images and their corresponding 3,000 single-channel digital labels, ensuring that each remote sensing surface water image in the dataset corresponds to its unique and precisely matched single-channel digital label.

[0042] As an implementation method of the embodiment of the present invention, in step S2, as Figure 2 、 3 As shown in the figure, the Remote Sensing Surface Water Body Extraction Network (RSWE-Net) extracts surface water bodies from remote sensing images using RSWE-Net. The process of extracting surface water bodies from remote sensing images is divided into three steps: extracting and integrating cross-scale features of remote sensing surface water images; extracting multi-level small-target remote sensing surface water body features; and extracting contextual semantic features of remote sensing surface water images. These three main steps correspond to the three modules of the RSWE-Net network model: the cross-scale adaptive feature integration encoding module; the multi-level small-target water body accuracy optimization module; and the contextual semantic attention decoding module.

[0043] Further, step S2 specifically includes:

[0044] Step 2.1 Extract cross-scale features of remote sensing surface water images and integrate them

[0045] The task of extracting and integrating cross-scale features of remote sensing surface water images is completed by the cross-scale adaptive feature integration encoding module of RSWE-Net.

[0046] The operation and construction process of the cross-scale adaptive feature integration coding module is as follows:

[0047] Step 2.1.1: Input the remote sensing surface water image into the cross-scale adaptive feature integration coding module.

[0048] Step 2.1.2: Use a convolution kernel of size 3×3 to perform convolution operation on the remote sensing surface water image, and then perform batch normalization (i.e. Figure 2BN operation in the , where BN is called BatchNorm) and activation function processing (i.e. Figure 2 The ReLU operation in is used to obtain the shallow feature map F1 of the first water body.

[0049] Step 2.1.3: Use the convolution kernel of size 3×3 to perform convolution operation on F1 to obtain the second shallow water feature map F2, and input F2 into the three branches including A1, A2 and A3 (see Figure 2 ).

[0050] Step 2.1.4: A hybrid multi-path pooling convolutional block (HMPCB) is designed and introduced in the A1 branch. In the A1 branch, HMPCB is used to extract cross-scale features of F2. The overall structure of the HMPCB in the A1 branch is as follows: Figure 4 As shown:

[0051] Input F2 into HMPCB, first, apply global average pooling (i.e. Figure 4 Global Avg Pool operation in), global maximum pooling (i.e. Figure 4 The Global Max Pool operation in ) and the convolution kernel of size 3×3 perform pooling and convolution operations on F2 respectively; then, the three results are spliced ​​in the channel dimension (i.e. Figure 4 Concat operation in the third water body) to obtain the shallow feature map F2_1; then use the convolution kernel with a size of 3×3 and a void rate of 3 to perform a convolution operation on F2_1; then use the Sigmoid activation function to obtain the weight ω of each channel; finally, the weight ω is multiplied by F2 element by element (i.e. Figure 4 The M operation in is performed to obtain the first cross-scale adaptive feature map F3.

[0052] The process of processing F2 with HMPCB to obtain F3 is shown in the following equations (1), (2) and (3):

[0053] F2_1=Concat(GAP(F2),GMP(F2),Conv3(F2)) (1)

[0054] ω=S(Conv3_3(F2_1)) (2)

[0055] F3=ω⊙F2 (3)

[0056] In the above formula, Concat() represents the concatenation operation on the channel dimension; GAP() represents the Global Avg Pool operation; GMP() represents the Global Max Pool operation; Conv3() represents the convolution operation of a convolution kernel of size 3×3; Conv3_3() represents the convolution operation of a convolution kernel of size 3×3 and a void rate of 3; S represents the Sigmoid activation function; ⊙ represents the element-by-element product.

[0057] Among them, the dilation rate: The ordinary convolution operation calculates the pixel value of the corresponding position in the output feature map by multiplying the convolution kernel with each pixel point of the input feature map and summing them. The dilation convolution inserts additional intervals (or "holes") in the convolution kernel. These intervals can increase the effective size of the convolution kernel without increasing the number of parameters. The size of this interval is called the "dilation rate" and is usually expressed as an integer. It determines the size of the interval in the convolution kernel.

[0058] Concat channel dimension splicing: refers to splicing two or more features with the same spatial dimensions (height and width) on the channel dimension to generate a new feature map. This new feature map has the same spatial dimensions as the input feature map, but the number of channels is the sum of the number of channels of all input feature maps.

[0059] Step 2.1.4: After concatenating F3 and F2 in the channel dimension, input them into HMPCB and process them in the same way as HMPCB on branch A1 to obtain the second cross-scale adaptive feature map F4.

[0060] Step 2.1.5: Input F4 into the A3 branch, concatenate F4 and F2 in the channel dimension, and then perform a 3×3 convolution operation to obtain the third cross-scale adaptive feature map F5.

[0061] Step 2.1.6: After concatenating F2, F3, F4, and F5 in the channel dimension, perform a 3×3 convolution operation to obtain the fourth cross-scale adaptive feature map F6.

[0062] Step 2.2 Extract the characteristics of multi-level small-target remote sensing surface water bodies

[0063] The task of extracting the characteristics of multi-level small-target remote sensing surface water bodies is completed by the multi-level small-target water body precision optimization module of RSWE-Net. The multi-level small-target water body precision optimization module consists of five parallel branches (B1, B2, B3 and B4). Among them, the overall structure of the multi-level small-target water body precision optimization module is as follows: Figure 5 As shown:

[0064] The operation and construction process of the multi-level small target water body precision optimization module is as follows:

[0065] Step 2.2.1: Input F6 into the B1 branch and perform convolution operation on F6 using a 1×1 convolution kernel and a 3×3 convolution kernel to obtain the first multi-level small target water body feature map F6_1.

[0066] Step 2.2.2: Input F6 into the B2 branch, and then use the pooling convolution normalization activation submodule PCNAB (Pooling Convolution Normalization Activation Block) to process F6. The overall structure of PCNAB is as follows: Figure 6 As shown:

[0067] In PCNAB, first, horizontal average pooling (i.e. Figure 6 X_Avg Pool operation in ) and vertical average pooling (i.e. Figure 6 The Y_Avg Pool operation in the pooling operation is performed on F6 to obtain the vertical and horizontal outputs; the vertical and horizontal outputs are then fused in the channel dimension (i.e. Figure 6 Concat operation in ); then use the convolution kernel of size 3×3 to perform convolution operation on the fusion result, and then perform batch normalization operation (i.e. Figure 6 The BN operation in ( ) is used to obtain the weight γ of each channel, and then the element-by-element product is performed with F6 to obtain the second multi-level small target water body feature map F6_2. The convolution kernel of size 3×3 is then used to perform a convolution operation on F6_2, and it is activated by the Sigmoid function. Finally, the channel dimension is fused with F6_2 to obtain the seventh multi-level small target water body feature map F7.

[0068] The process of using PCNAB to process F6 to obtain F7 is shown in the following equations (4), (5) and (6):

[0069] γ=BN(Conv3(Concat(XAP(F6),YAP(F6)))) (4)

[0070] F6_3=F6⊙γ (5)

[0071] F7=Concat(F6,S(Conv3(F6_3))) (6)

[0072] Among them, Concat() represents the concatenation operation on the channel dimension; XAP() represents the X_Avg Pool operation; YMP() represents the Y_Avg Pool operation; Conv3() represents the convolution operation of the convolution kernel with a size of 3×3; BN() represents the batch normalization operation; S represents the Sigmoid activation function; ⊙ represents the element-by-element product.

[0073] Step 2.2.3: Input F6 into branch B3 and perform convolution operations on F6 using a 1×1 convolution kernel, a 5×5 convolution kernel, and a 3×3 dilated convolution with a dilation ratio of 5. This yields the third multi-level small target water feature map F6_3. F7 is fused with F6_3 in the channel dimension to yield the fourth multi-level small target water feature map F6_4.

[0074] Step 2.2.4: Input F6 into the B4 branch and perform convolution operations on F6 using a 1×1 convolution kernel, a 9×9 convolution kernel, and a 3×3 dilated convolution with a dilation ratio of 9 to obtain the fifth multi-level small target water body feature map F6_5. Then, F7 and F6_5 are fused in the channel dimension to obtain the sixth multi-level small target water body feature map F6_6.

[0075] Step 2.2.5: F6_1, F6_4 and F6_6 are fused in the channel dimension and activated using the Sigmoid activation function to obtain the eighth multi-level small target water body feature map F8.

[0076] Step 2.3 Extracting contextual semantic features of remote sensing surface water images

[0077] The task of extracting contextual semantic features of remote sensing surface water images is completed by the contextual semantic attention decoding module of RSWE-Net.

[0078] The operation and construction process of the contextual semantic attention decoding module is as follows:

[0079] Step 2.3.1: Input F8 into the contextual semantic attention decoding module, use bilinear interpolation to resize F7, and obtain the first contextual semantic feature map F9.

[0080] Step 2.3.2: Use a 1×1 convolution kernel to convolve F9 to obtain the second contextual semantic feature map F10, and then input F10 into the four branches C1, C2, C3 and C4.

[0081] Step 2.3.3: Input F10 into C1 and perform the maximum pooling operation on F10 (i.e. Figure 2The MaxPool operation shown in ) and the convolution operation using a convolution kernel of size 1×1 are performed to obtain the third contextual semantic feature map F10_1.

[0082] Step 2.3.4: Input F10 into C2. First, perform a convolution operation on F10 using a convolution kernel of size 1×1. Then, perform a further convolution operation using a dilated convolution kernel of size 1×1 and a dilation rate of 3 to obtain the fourth contextual semantic feature map F10_2.

[0083] Step 2.3.5: Input F10 into C3. First, perform a convolution operation on F10 using a convolution kernel of size 1×1. Then, perform a further convolution operation using a dilated convolution kernel of size 1×1 and a dilation rate of 5 to obtain the fifth contextual semantic feature map F10_3.

[0084] Step 2.3.6: Input F10 into C4 and perform average pooling operation on F10 (i.e. Figure 2 A convolution operation is performed using a convolution kernel of size 1×1 to obtain the sixth contextual semantic feature map F10_4.

[0085] Step 2.3.7: First perform bilinear interpolation on F10_1, F10_2, F10_3, and F10_4, and then fuse them with F2 in the channel dimension to obtain the seventh contextual semantic feature map F11.

[0086] Step 2.3.8: Use a 1×1 convolution to convolve F11 and adjust the number of channels of F11 to obtain a surface water semantic segmentation image. In the surface water semantic segmentation image, white represents surface water and black represents background.

[0087] In the experiment of the present invention, a remotely sensed surface water image with a size of 512×512 pixels and 3 channels was input into the cross-scale adaptive feature integration coding module. A convolution operation was performed on the remotely sensed surface water image using a 3×3 convolution kernel, followed by normalization and ReLU activation function processing to obtain a first shallow water feature map F1 with a size of 256×256 pixels and 64 channels. A convolution operation was performed on F1 using a 3×3 convolution kernel to obtain a second shallow water feature map F2 with a size of 128×128 pixels and 128 channels. F2 is fed into three branches, A1, A2, and A3. In branch A1, HMPCB processing is performed to obtain the first cross-scale adaptive feature map F3, which is 128×128 pixels and has 128 channels. F3 is fed into branch A2, where it is concatenated with F2 along the channel dimension and then fed into HMPCB. The same processing is performed as in branch A1, resulting in the second cross-scale adaptive feature map F4, which is 128×128 pixels and has 128 channels. F4 is fed into branch A3, where it is concatenated with F2 along the channel dimension and then subjected to a 3×3 convolution operation to obtain the third cross-scale adaptive feature map F5, which is 128×128 pixels and has 128 channels. After concatenating F2, F3, F4, and F5 in the channel dimension, a 3×3 convolution operation is performed to obtain the fourth cross-scale adaptive feature map F6, which has a size of 128×128 pixels and 128 channels.

[0088] F6 is fed into the B1 branch of the multi-level small-target water body precision optimization module. It is convolved with a 1×1 convolution kernel and a 3×3 convolution kernel to produce the first multi-level small-target water body feature map F6_1, which is 128×128 pixels and 128 channels. F6 is fed into the B2 branch and processed using the pooling, convolution, normalization, and activation submodule PCNAB to produce the seventh multi-level small-target water body feature map F7, which is 128×128 pixels and 128 channels. Input F6 into the B3 branch, and use the convolution kernel of size 1×1, the convolution kernel of size 5×5, and the dilated convolution of size 3×3 and dilated rate 5 to perform convolution operation on F6 to obtain the third multi-level small target water body feature map F6_3, the size of F6_3 is 128×128 pixels and 128 channels. F7 is fused with F6_3 in the channel dimension to obtain the fourth multi-level small target water body feature map F6_4, the size of F6_4 is 128×128 pixels and 128 channels. F6 is input into branch B4, and convolution operations are performed on F6 using a 1×1 convolution kernel, a 9×9 convolution kernel, and a 3×3 dilated convolution with a dilation ratio of 9. This yields the fifth multi-level small target water body feature map F6_5, which is 128×128 pixels and has 128 channels. F7 is then fused with F6_5 in the channel dimension to yield the sixth multi-level small target water body feature map F6_6, which is 128×128 pixels and has 128 channels. F6_1, F6_4, and F6_6 are fused in the channel dimension and activated using the Sigmoid activation function to yield the eighth multi-level small target water body feature map F8, which is 128×128 pixels and has 128 channels.

[0089] F8 is input into the contextual semantic attention decoding module. First, bilinear interpolation is used to adjust the size of F8 to obtain the first contextual semantic feature map F9, which has a size of 256×256 pixels and 64 channels. Then, a convolution operation is performed on F9 with a convolution kernel of size 1×1 to obtain the second contextual semantic feature map F10, which has a size of 256×256 pixels and 64 channels. F10 is then input into four branches: C1, C2, C3, and C4. Input F10 into C1, perform a maximum pooling operation on F10 and a convolution operation using a convolution kernel of size 1×1 to obtain the third contextual semantic feature map F10_1, with a size of 256×256 pixels and 64 channels; input F10 into C2, first perform a convolution operation on F10 using a convolution kernel of size 1×1, and then perform a further convolution operation using a dilated convolution kernel of size 1×1 and a dilation rate of 3 to obtain the fourth contextual semantic feature map F10_2, with a size of 256×256 pixels and 64 channels; F10 is input to C3, where it is first convolved with a 1×1 kernel and then further convolved with a 1×1 dilated kernel with a dilation ratio of 5. This yields the fifth contextual semantic feature map F10_3, which has a size of 256×256 pixels and 64 channels. F10 is then input to C4, where it is average pooled and convolved with a 1×1 kernel to yield the sixth contextual semantic feature map F10_4, which has a size of 256×256 pixels and 64 channels. F10_1, F10_2, F10_3, and F10_4 are bilinearly interpolated and then fused with F2 in the channel dimension to yield the seventh contextual semantic feature map F11, which has a size of 512×512 pixels and 32 channels. A convolution operation is performed on F11 using a convolution of size 1×1, and the number of channels of F11 is adjusted to obtain a semantic segmentation image of the surface water body. The size of the semantic segmentation image of the surface water body is 512×512 pixels and 1 channel.

[0090] As one implementation of this embodiment of the present invention, in step S3, the remote sensing surface water dataset is divided into a training set, a validation set, and a test set according to a certain ratio. The training set, validation set, and test set are mutually exclusive. The training set accounts for 70% of the total dataset, namely 2,100 remote sensing surface water images. The validation set and test set each account for 15% of the total dataset, namely 450 remote sensing surface water images each.

[0091] As one implementation of an embodiment of the present invention, step S4 specifically involves initializing all neural network parameters and setting model-related hyperparameters, such as the number of training rounds, batch size, optimizer selection, learning rate, and total number of iterations. After initializing the parameters, the training and validation data sets are divided into multiple batches. Each batch of training data is input into the RSWE-Net remote sensing surface water extraction network for training, and the training loss value for that batch is obtained. After completing one round of training for all batches of the training data, the validation data set is input into the RSWE-Net remote sensing surface water extraction network batch by batch, and the corresponding batch loss value, batch_loss, is obtained. The validation loss value is primarily used to monitor overfitting and adjust the training strategy, such as terminating training early or adjusting the learning rate. During training and validation, the RSWE-Net remote sensing surface water extraction network automatically learns and adjusts parameters based on each loss and batch_loss. Training of the RSWE-Net remote sensing surface water extraction network ends when the batch_loss value converges after one or more rounds of training.

[0092] As an implementation method of an embodiment of the present invention, in step S5, the test set is input into the trained RSWE-Net, and the surface water body extraction result is output, that is, the surface water body semantic segmentation image, which removes background interference and only retains the water body.

[0093] In the cross-scale adaptive feature integration coding module of RSWE-Net, an embodiment of the present invention designs a hybrid multi-path pooling convolution sub-module HMPCB. HMPCB uses global average pooling, global maximum pooling and void convolution kernel to calculate the channel weights of each input image, and then multiplies the weights by the shallow feature map element by element to obtain a cross-scale adaptive feature map. HMPCB is connected in parallel with ordinary convolution to jointly extract and integrate feature maps with feature information of different scales.

[0094] In the multi-level small target water body precision optimization module of RSWE-Net, an embodiment of the present invention designs four parallel branches, namely B1, B2, B3 and B4. A pooling convolution normalization activation submodule PCNAB is designed in the multi-level small target water body precision optimization module. PCNAB extracts features from the horizontal and vertical directions, and then performs normalization and activation function processing to obtain a multi-level small target water body feature map. PCNAB increases the nonlinear capability of the model. Then, PCNAB is combined with ordinary convolution, void convolution and activation function to extract features of small target water bodies in remote sensing images, solving the problem that the detailed information of extremely small water bodies is not clear enough, making extraction difficult, and improving the accuracy of water body detection.

[0095] In the contextual semantic attention decoding module of RSWE-Net, this embodiment of the present invention designs four branches consisting of maximum pooling, average pooling, dilated convolution, and normal convolution. The extraction results of these four branches are fused along the channel dimension and then fused with the shallow feature map in the cross-scale adaptive feature integration encoding module along the channel dimension to obtain a contextual semantic feature map.

[0096] The embodiments of the present invention can integrate feature maps of feature information of different scales to solve the problem that water bodies of different scales present large differences in features on remote sensing images; the detection accuracy of small target water bodies is higher, which solves the problem that the detailed information of water bodies in extremely small areas is not clear enough due to the continuous improvement of the resolution of remote sensing images, making extraction difficult; the semantic information attention ability is strong, which reduces the influence of the complex correlation between water bodies in remote sensing images and the surrounding environment, and focuses on the extraction of contextual semantic information of surface water bodies in remote sensing images during the extraction process.

[0097] Example 2:

[0098] An embodiment of the present invention further provides a device for extracting surface water from a remote sensing image, comprising:

[0099] Acquisition module, used to obtain remote sensing surface water datasets;

[0100] A construction module is used to construct a remote sensing surface water extraction network RSWE-Net based on a remote sensing surface water dataset. The RSWE-Net includes: a cross-scale adaptive feature integration encoding module; a multi-level small target water accuracy optimization module; and a contextual semantic attention decoding module.

[0101] The partitioning module is used to partition the remote sensing surface water dataset into training and test sets;

[0102] The training module is used to train the remote sensing surface water extraction network RSWE-Net based on the training set;

[0103] The extraction module is used to input the test set into the trained remote sensing surface water extraction network RSWE-Net to extract surface water bodies in remote sensing images.

[0104] Example 3:

[0105] An embodiment of the present invention also provides a remote sensing image surface water body extraction system, comprising: a memory and a processor, wherein the memory stores a computer program run by the processor, and the computer program executes a remote sensing image surface water body extraction method when run by the processor.

[0106] Example 4:

[0107] An embodiment of the present invention further provides a storage medium having a computer program stored thereon. When the computer program is run, the method for extracting surface water bodies from remote sensing images is executed.

[0108] The embodiments described above are merely descriptions of preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Without departing from the spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by persons skilled in the art should fall within the scope of protection defined by the claims of the present invention.

Claims

1. A method for extracting surface water from remote sensing images, characterized in that: include: Step S1, obtaining a remote sensing surface water dataset; Step S2: constructing a remote sensing surface water extraction network RSWE-Net based on the remote sensing surface water dataset; wherein the remote sensing surface water extraction network RSWE-Net includes: a cross-scale adaptive feature integration encoding module; a multi-level small target water body precision optimization module; and a contextual semantic attention decoding module; Step S3, dividing the remote sensing surface water dataset to obtain a training set and a test set; Step S4, training the remote sensing surface water extraction network RSWE-Net according to the training set; Step S5: input the test set into the trained remote sensing surface water extraction network RSWE-Net to extract surface water bodies in the remote sensing image; The cross-scale adaptive feature integration coding module includes: a hybrid multi-path pooling convolution sub-module, which combines the pooling residual mechanism with ordinary convolution in parallel to extract cross-scale adaptive feature maps; The multi-level small-target water body precision optimization module includes: a pooled convolution normalized activation submodule, which combines horizontal and vertical extraction mechanisms with different types of dilated convolutions in parallel to extract multi-level small-target water body feature maps with different receptive fields; The contextual semantic attention decoding module learns and synthesizes different feature information through four different branches consisting of maximum pooling, average pooling, void convolution and ordinary convolution, and simultaneously splices the shallow feature map with the contextual semantic feature map in the channel dimension.

2. A remote sensing image surface water extraction device for implementing the remote sensing image surface water extraction method according to claim 1, characterized in that: include: Acquisition module, used to obtain remote sensing surface water datasets; A construction module is used to construct a remote sensing surface water extraction network RSWE-Net based on a remote sensing surface water dataset. The RSWE-Net includes: a cross-scale adaptive feature integration encoding module; a multi-level small target water accuracy optimization module; and a contextual semantic attention decoding module. The partitioning module is used to partition the remote sensing surface water dataset into training and test sets; The training module is used to train the remote sensing surface water extraction network RSWE-Net based on the training set; The extraction module is used to input the test set into the trained remote sensing surface water extraction network RSWE-Net to extract surface water bodies in remote sensing images.

3. A remote sensing image surface water extraction system, characterized in that: include: A memory and a processor, wherein the memory stores a computer program executed by the processor, and when the computer program is executed by the processor, the method for extracting surface water bodies from remote sensing images according to claim 1 is executed.

4. A storage medium, characterized in that The storage medium stores a computer program, which executes the method for extracting surface water bodies from remote sensing images according to claim 1 when running.

Images