Typhoon period wave climbing identification method and device

Abstract

The embodiment of the application discloses a typhoon period wave climbing identification method and device, comprising classifying the collected typhoon period overwave pictures and normal period wave climbing pictures; processing the normal period wave climbing pictures to obtain semantic segmentation binary graphs of wave regions and first background regions; processing the typhoon period overwave pictures to obtain semantic segmentation binary graphs of overwave regions and second background regions, and identifying a dike main body in the second background region; calibrating and measuring multiple cross-section information of the dike main body to obtain calibration information, combining the wave regions with the calibration information to obtain wave climbing points, and combining the overwave regions with the calibration information to obtain overwave highest points. Through the improved algorithm, the collected typhoon period overwave pictures and normal period wave climbing pictures are identified and classified, the two kinds of pictures are processed respectively, the dike main body in the pictures is calibrated and measured, and then the wave climbing points and the overwave highest points are extracted.

Description

Technical Field

[0001] This invention relates to the field of artificial intelligence technology, specifically to a method and apparatus for identifying wave rise during typhoon season. Background Technology

[0002] Currently, seawalls are crucial hydraulic structures for resisting wave erosion and protecting buildings behind them. Wave run-up is the vertical height a wave climbs along a slope, also known as wave ramp height. When a wave reaches the shore, most of its energy is lost as it breaks up, while some energy is converted into potential energy as the wave climbs up the sand. In coastal engineering, wave run-up is a major factor influencing beach dune erosion and deposition. Determining wave height within the wave run-up range has a significant impact on seawall stability and the safety of people and property along the coast.

[0003] Because typhoon seasons are infrequent, levees remain in a stable phase of high and low tides for extended periods. However, when typhoons arrive, they are often accompanied by torrential rains and massive storm surges, causing water levels to rise significantly and resulting in overtopping. Since overtopping data is often scarce, while high and low tide data is abundant during calm periods, using a single model for identification would lead to uneven distribution of positive and negative samples. Summary of the Invention

[0004] To address the aforementioned shortcomings, this invention discloses a method and apparatus for identifying wave run-up during typhoon seasons, which can accurately identify wave run-up.

[0005] The first aspect of this invention discloses a method for identifying wave run-up during typhoon season, comprising:

[0006] Acquire images of waves crossing during typhoon seasons and waves climbing during normal seasons, and classify the collected images of waves crossing during typhoon seasons and waves climbing during normal seasons.

[0007] The image of the wave rise during the normal period is processed to obtain a semantic segmentation binary image of the wave region and the first background region;

[0008] The images of typhoon-era waves are processed to obtain semantic segmentation binary maps of the wave-over region and the second background region, and the main body of the dam in the second background region is identified.

[0009] The main body of the dam is divided into multiple cross-sections for calibration measurement to obtain calibration information. The wave area is combined with the calibration information to obtain the wave run-up point, and the wave overtopping area is combined with the calibration information to obtain the highest wave overtopping point.

[0010] As an optional implementation, in the first aspect of the present invention, calibration information is obtained by calibrating the main body of the dam by dividing it into multiple cross-sectional information, including:

[0011] The main body of the dam was divided into multiple sections, and the height and length of the dam were measured for each section.

[0012] RTK data of the main body of the dam was collected, and interpolation was performed on the RTK data to obtain the pixel positions of different survey lines.

[0013] Obtain the dam height corresponding to each cross section and the pixel corresponding to the elevation of that dam height, thus forming a mapping relationship between dam height and pixels.

[0014] As an optional implementation, in a first aspect of the present invention, the step of combining the wave region with the calibration information to obtain the wave climb point, and combining the wave overshoot region with the calibration information to obtain the highest wave overshoot point, includes:

[0015] Extract the first most convex point of the wave region and the second most convex point of the overpass region;

[0016] The calibration information of one cross section corresponding to the first or second most convex point is selected from multiple cross section information, and the wave climb point or the highest point of wave overpass is obtained based on the calibration information.

[0017] As an optional implementation, in the first aspect of the present invention,

[0018] The collected images of wave overtopping during typhoon seasons and wave rise during normal seasons were categorized, including:

[0019] A classification model was built using the improved EfficientNetV2 algorithm;

[0020] The collected images of typhoon-period wave overpasses and normal-period wave rises are input into the classification model for classification, and the classified images of typhoon-period wave overpasses and normal-period wave rises are output.

[0021] As an optional implementation, in the first aspect of the present invention, the method further includes: marking the identified dam body with a rectangular frame.

[0022] As an optional implementation, in the first aspect of the present invention, it further includes:

[0023] Obtain the overlapping area between the wave-crossing area and the main body of the dam, and calculate the area ratio of the overlapping area on the main body of the dam;

[0024] When the area ratio exceeds a set threshold, the dwell time of the current overtopping area on the main body of the embankment is obtained. If the dwell time exceeds the set time, it is determined to be a severe overtopping.

[0025] As an optional implementation, in the first aspect of the present invention, the wave rise image during the normal period is processed by the improved PSPNet algorithm, the typhoon period image is processed by the improved UET algorithm, and the dam body in the second background area is identified by the improved YOLOv3 algorithm.

[0026] A second aspect of this invention discloses a wave run-up identification device during typhoon season, comprising:

[0027] Image collection module: used to acquire images of waves crossing during typhoon season and waves climbing during normal season, and to classify the collected images of waves crossing during typhoon season and waves climbing during normal season;

[0028] The first processing module is used to process the wave rise image during the normal period to obtain a semantic segmentation binary image of the wave region and the first background region.

[0029] The second processing module is used to process the images of typhoon-era waves to obtain semantic segmentation binary images of the wave-overlapping area and the second background area, and to identify the main body of the dam in the second background area.

[0030] High point acquisition module: used to calibrate and measure the main body of the dam into multiple cross sections to obtain calibration information, combine the wave area with the calibration information to obtain the wave climb point, and combine the wave overpass area with the calibration information to obtain the highest point of the wave overpass.

[0031] As an optional implementation, in a second aspect of the present invention, calibration information is obtained by calibrating the main body of the dam by dividing it into multiple cross-sectional information, including:

[0032] The main body of the dam was divided into multiple sections, and the height and length of the dam were measured for each section.

[0033] RTK data of the main body of the dam was collected, and interpolation was performed on the RTK data to obtain the pixel positions of different survey lines.

[0034] Obtain the dam height corresponding to each cross section and the pixel corresponding to the elevation of that dam height, thus forming a mapping relationship between dam height and pixels.

[0035] As an optional implementation, in a second aspect of the present invention, the step of combining the wave region with the calibration information to obtain the wave climb point, and combining the overtopping region with the calibration information to obtain the highest overtopping point, includes:

[0036] Extract the first most convex point of the wave region and the second most convex point of the overpass region;

[0037] The calibration information of one cross section corresponding to the first or second most convex point is selected from multiple cross section information, and the wave climb point or the highest point of wave overpass is obtained based on the calibration information.

[0038] As an optional implementation, in a second aspect of the present invention,

[0039] The collected images of wave overtopping during typhoon seasons and wave rise during normal seasons were categorized, including:

[0040] A classification model was built using the improved EfficientNetV2 algorithm;

[0041] The collected images of typhoon-period wave overpasses and normal-period wave rises are input into the classification model for classification, and the classified images of typhoon-period wave overpasses and normal-period wave rises are output.

[0042] As an optional implementation, in the second aspect of the present invention, the method further includes: marking the identified dam body with a rectangular frame.

[0043] As an optional implementation, in the second aspect of the present invention, it further includes:

[0044] Obtain the overlapping area between the wave-crossing area and the main body of the dam, and calculate the area ratio of the overlapping area on the main body of the dam;

[0045] When the area ratio exceeds a set threshold, the dwell time of the current overtopping area on the main body of the embankment is obtained. If the dwell time exceeds the set time, it is determined to be a severe overtopping.

[0046] As an optional implementation, in the second aspect of the present invention, the wave rise image during the normal period is processed by the improved PSPNet algorithm, the typhoon period image is processed by the improved UET algorithm, and the dam body in the second background area is identified by the improved YOLOv3 algorithm.

[0047] A third aspect of the present invention discloses an electronic device, comprising: a memory storing executable program code; a processor coupled to the memory; the processor calling the executable program code stored in the memory to execute the typhoon wave rise identification method disclosed in the first aspect of the present invention.

[0048] The fourth aspect of this invention discloses a computer-readable storage medium storing a computer program, wherein the computer program causes a computer to execute the typhoon wave rise identification method disclosed in the first aspect of this invention.

[0049] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0050] In this embodiment of the invention, an improved algorithm is used to identify and classify the collected images of wave overtopping during typhoon season and wave rise during normal season. The two types of images are processed separately, and the main body of the dam in the images is calibrated and measured to extract the wave rise point and the highest point of wave overtopping. Attached Figure Description

[0051] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0052] Figure 1 This is a flowchart illustrating a method for identifying wave run-up during typhoon season disclosed in an embodiment of the present invention.

[0053] Figure 2 This is a flowchart illustrating another method for identifying wave run-up during typhoon season disclosed in an embodiment of the present invention.

[0054] Figure 3 This is a diagram showing the improved network structure of the EfficientNetV2 algorithm disclosed in an embodiment of the present invention;

[0055] Figure 4 This is a schematic diagram of the convolution principle of the Fused-MBconv module disclosed in an embodiment of the present invention;

[0056] Figure 5 This is a schematic diagram of the convolution principle of the MBconv module disclosed in an embodiment of the present invention;

[0057] Figure 6 This is a schematic diagram of the structure of a wave run-up identification device during typhoon season provided in an embodiment of the present invention;

[0058] Figure 7 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0059] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and 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.

[0060] It should be noted that the terms "first," "second," "third," "fourth," etc., in the specification and claims of this invention are used to distinguish different objects, not to describe a specific order. The terms "comprising" and "having," and any variations thereof, in the embodiments of this invention are intended to cover non-exclusive inclusion. Exemplarily, a process, method, system, product, or device that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to these processes, methods, products, or devices.

[0061] This invention discloses a method, device, electronic device, and storage medium for identifying wave run-up during typhoon season. The method uses an improved algorithm to identify and classify collected wave run-up images from typhoon season and wave run-up images from normal season. The two types of images are processed separately, and the main body of the embankment in the images is calibrated and measured to extract the wave run-up point and the highest point of the wave run-up.

[0062] Example 1

[0063] Please see Figure 1 , Figure 1 This is a flowchart illustrating a method for identifying wave rise during typhoon season, as disclosed in an embodiment of the present invention. The execution entity of the method described in this embodiment is an execution entity composed of software and / or hardware. This execution entity can receive relevant information via wired or / or wireless means and can send certain instructions. It may also have certain processing and storage functions. This execution entity can control multiple devices, such as remote physical servers or cloud servers and related software, or local hosts or servers and related software that perform related operations on devices located in a certain location. In some scenarios, it can also control multiple storage devices, which may be placed in the same location as the devices or in different locations. Figure 1 As shown, the wave run-up identification method during typhoon season includes the following steps:

[0064] 101. Obtain images of wave overtopping during typhoon season and wave rise during normal season, and classify the collected images of wave overtopping during typhoon season and wave rise during normal season.

[0065] The typhoon season is the period during which meteorological observatories monitor and indicate the occurrence of tropical cyclones. In this embodiment, the normal period is defined as the non-typhoon season. The embodiment first collects images of wave overtopping during the typhoon season and wave rise during the normal period. The wave overtopping images during the typhoon season are designated as label 0, and the wave rise images during the normal period are designated as label 1, constructing an image classification dataset. An algorithm is then used to build a classification model to classify wave overtopping during the typhoon season and wave rise during the normal period.

[0066] Specifically, the embodiment constructs a classification model using the improved EfficientNetV2 algorithm; the collected images of typhoon-period wave overtopping and normal-period wave rise are input into the classification model for classification, and the classified images of typhoon-period wave overtopping and normal-period wave rise are output.

[0067] See the improved EfficientNetV2 algorithm. Figure 3 It introduces the CBAM module, inserting it after the MBConv6 layer to extract deep features. It also replaces the BN layer in the 3x3 convolution + silu activation function + BN layer of stage0 conv with a GN layer. Specifically, the BN layer in the 3x3 conv of stage0 is replaced with a GN layer to address the poor performance of BN layers with small batch sizes. In stages1-3, the numbers 1 and 4 after the module name represent expansion ratios, and k3x3 represents the convolution kernel size. The Fused-MBconv module is as follows: Figure 4 As shown, in Stages 4-7, 4 and 6 after the module name represent the expansion ratio, k3*3 represents the convolution kernel size, and 0.25 is the input of the first fully connected layer node in the SE module, which is 0.25 of the channels of the feature matrix of that MBconv module. MBconv is as follows... Figure 5 As shown.

[0068] 102. Process the image of the wave rise during the normal period to obtain a semantic segmentation binary image of the wave region and the first background region.

[0069] 103. Process the images of typhoon-era waves to obtain semantic segmentation binary images of the wave-over region and the second background region, and identify the main body of the dam in the second background region.

[0070] In this embodiment, the improved PSPNet algorithm is used to process the wave rise image during the normal period, the improved UET algorithm is used to process the typhoon period image, and the improved YOLOv3 algorithm is used to identify the main body of the dam in the second background region.

[0071] The improved PSPNet algorithm replaces the ResNet101 backbone feature extraction network with DenseNet to enhance gradient backpropagation. To preserve negative features of the image, the ReLU activation function is replaced with the Swish activation function, improving segmentation performance. The Swish activation function has the properties of being unbounded at the upper limit but bounded at the lower limit, smooth, and non-monotonic. Its linear representation is similar to ReLU, and it can collect negative features: f(x) = x·sigmoid(βx), where β is a common-sense or trainable parameter. Swish has the characteristics of being bounded at the lower limit, smooth, and non-monotonic. Swish outperforms ReLU in deep models. The pooling scale of the PPM module is modified to 1*1, 2*2, 4*4, and 6*6 to increase the receptive field of features. An attention mechanism, CBAM, is added after pooling to enhance feature extraction capabilities. Specifically, the input image is processed by a DenseNet to extract features, and the extracted features (specifically, stride = 8) are used as input to the subsequent pyramid pooling module. In the pyramid pooling module, a feature pyramid with a depth of 4 is constructed. Features of different depths are based on input features, which are feature maps obtained from the original image after convolution and pooling operations. Convolution and pooling are common operations in deep learning. In this example, pooling is used to increase the receptive field. The receptive field is the size of the region mapped by a pixel on the feature map output by each layer in the convolutional neural network onto the input image; that is, a point on the feature map corresponds to a region on the input image. The pooling scale can be adjusted. The feature size after pooling is modified to 1*1, 2*2, 4*4, and 6*6 to increase the receptive field. Then, a 1*1 convolutional layer is used to reduce the feature dimension to 1 / 4 of the original, while an attention mechanism (CBAM) module is added. Finally, these pyramid features are directly upsampled to the same size as the input features and then merged with the input features, i.e., the concat operation, to obtain the final output feature map. The feature merging process is actually the process of fusing the detailed features (i.e., shallow features) and global features (deep features, i.e., contextual information) of the target. Here, because the receptive field of the feature layer output by the feature extraction network is large enough, there is sufficient global information.

[0072] The improved UET algorithm replaces the conv module in UET with a depthwise over-parameterized convolution, adding an extra depthwise convolution operation to a regular convolutional layer to form an over-parameterized convolutional layer. Using DO-Conv not only accelerates the network training process but also achieves better results than traditional convolutional layers in various computer vision tasks. Depthwise over-parameterized convolution adds an over-parameterized component to the CNV convolution. The depthwise convolution operator ° is applied to the depthwise convolution kernel D and the input feature P, generating the transformed feature P·=D°P. Then, the regular convolution operation * is applied to the regular convolution kernel W and the feature P1, generating the transformed feature O=W*P. · Therefore, the output of the depthwise overparameterized convolution can be represented as 0 = W * (D°P), where MN is the spatial dimension of the input vector, Cin is the number of channels in the input vector, Cout is the number of channels in the output, and dmul = M * N is the depth multiplier of the depthwise convolution. At the same time, a CBAM module is added after the last two Do-conv modules of downsampling to improve the feature extraction capability.

[0073] 104. The main body of the dam is divided into multiple cross-sectional information for calibration measurement to obtain calibration information. The wave area is combined with the calibration information to obtain the wave rise point, and the wave overpass area is combined with the calibration information to obtain the highest wave overpass point.

[0074] Example 2

[0075] Please see Figure 2 , Figure 2 This is a flowchart illustrating another method for identifying wave run-up during typhoon season disclosed in an embodiment of the present invention. Figure 2 As shown, the wave run-up identification method during typhoon season includes:

[0076] 201. Obtain images of wave overtopping during typhoon season and wave rise during normal season, and classify the collected images of wave overtopping during typhoon season and wave rise during normal season.

[0077] 202. Process the image of the wave rise during the normal period to obtain a semantic segmentation binary image of the wave region and the first background region.

[0078] 203. Process the typhoon-era wave overtopping images to obtain semantic segmentation binary maps of the overtopping area and the second background area, and identify the main body of the dam in the second background area. In this embodiment, the identified dam main body is labeled. For example, a rectangular bounding box can be used to label the dam main body. In other examples, other forms of labeling can also be used.

[0079] More specifically, the embodiment obtains the overlapping area between the overtopping area and the main body of the dam, and calculates the area ratio of the overlapping area on the main body of the dam; when the area ratio exceeds a set threshold, the dwell time of the current overtopping area on the main body of the dam is obtained; if the dwell time exceeds the set time, it is determined to be a severe overtopping.

[0080] 204. Divide the main body of the dam into multiple sections and measure the dam height and dam length of each section.

[0081] 205. Collect RTK data of the main body of the dam, and perform interpolation and sorting based on the RTK data to obtain the pixel points of different survey lines.

[0082] 206. Obtain the dam height corresponding to each cross section and the pixel corresponding to the elevation of the dam height, and form a mapping relationship between dam height and pixels.

[0083] 207. Extract the first most convex point of the wave region and the second most convex point of the overpass region.

[0084] 208. Select the calibration information of one section information corresponding to the first or second most convex point from multiple section information, and obtain the wave climb point or the highest point of wave overpass based on the calibration information.

[0085] Example 3

[0086] Please see Figure 6 , Figure 6 This is a schematic diagram of the structure of a wave run-up identification device during typhoon season disclosed in an embodiment of the present invention. Figure 6 As shown, the typhoon wave rise recognition device may include an image collection module 601, a first processing module 602, a second processing module 603, and a high point acquisition module 604. The image collection module 601 is used to acquire images of wave rise during typhoon periods and wave rise during normal periods, and to classify the collected images. The first processing module 602 is used to process the wave rise images during normal periods to obtain semantic segmentation binary images of the wave region and a first background region. The second processing module 603 is used to process the wave rise images during typhoon periods to obtain semantic segmentation binary images of the wave rise region and the second background region, and to identify the dam body in the second background region. The high point acquisition module 604 is used to calibrate and measure the dam body into multiple cross-sectional information to obtain calibration information, combine the wave region with the calibration information to obtain the wave rise point, and combine the wave rise region with the calibration information to obtain the highest point of the wave rise.

[0087] In the image collection module 601 of this embodiment, a classification model is constructed using an improved EfficientNetV2 algorithm. Collected images of typhoon-period wave overtopping and normal-period wave rise are input into the classification model for classification, and the classified images of typhoon-period wave overtopping and normal-period wave rise are output. In the first processing module 602, the normal-period wave rise images are processed using an improved PSPNet algorithm. In the second processing module 603, the typhoon-period images are processed using an improved UET algorithm, and the dam body in the second background region is identified using an improved YOLOv3 algorithm. In the high-point acquisition module 604, specifically, the dam body is divided into multiple sections, and the dam height and length of each section are measured. RTK data of the dam body is collected, and interpolation is performed based on the RTK data to obtain the pixel positions of different survey lines. The dam height corresponding to each section and the pixels corresponding to the elevation of that dam height are obtained, forming a mapping relationship between dam height and pixels. Extract the first most convex point of the wave region and the second most convex point of the wave overpass region; select the calibration information of one cross-sectional information corresponding to the first or second most convex point from multiple cross-sectional information, and obtain the wave climb point or the highest point of the wave overpass based on the calibration information.

[0088] Example 4

[0089] Please see Figure 7 , Figure 7 This is a schematic diagram of the structure of an electronic device disclosed in an embodiment of the present invention. The electronic device can be a computer, a server, etc. Of course, in certain cases, it can also be a mobile phone, tablet computer, monitoring terminal, or other smart device, as well as an image acquisition device with processing capabilities. Figure 7 As shown, the electronic device may include:

[0090] Memory 701 storing executable program code;

[0091] Processor 702 coupled to memory 701;

[0092] The processor 702 calls the executable program code stored in the memory 701 to execute some or all of the steps in the typhoon wave rise identification method in Embodiment 1.

[0093] This invention discloses a computer-readable storage medium storing a computer program that causes a computer to perform some or all of the steps in the typhoon wave rise identification method of Embodiment 1.

[0094] This invention also discloses a computer program product, wherein when the computer program product is run on a computer, the computer executes some or all of the steps in the typhoon wave rise identification method in Embodiment 1.

[0095] This invention also discloses an application publishing platform, which is used to publish computer program products. When the computer program products are run on a computer, the computer executes some or all of the steps in the typhoon wave rise identification method in Embodiment 1.

[0096] In various embodiments of the present invention, it should be understood that the sequence number of each process does not necessarily imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0097] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they can be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0098] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0099] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-accessible memory. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a memory and includes several requests to cause a computer device (which can be a personal computer, server, or network device, specifically a processor in the computer device) to execute some or all of the steps of the methods described in the various embodiments of the present invention.

[0100] In the embodiments provided by this invention, it should be understood that "B corresponding to A" means that B is associated with A, and B can be determined based on A. However, it should also be understood that determining B based on A does not mean determining B solely based on A; B can also be determined based on A and / or other information.

[0101] Those skilled in the art will understand that some or all of the steps in the various methods of the embodiments described can be implemented by a program instructing related hardware. This program can be stored in a computer-readable storage medium, including read-only memory (ROM), random access memory (RAM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), one-time programmable read-only memory (OTPROM), electrically-erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc storage, disk storage, magnetic tape storage, or any other computer-readable medium capable of carrying or storing data.

[0102] The above provides a detailed description of the typhoon wave rise identification method, device, electronic equipment, and storage medium disclosed in the embodiments of the present invention. Specific examples have been used to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for identifying wave run-up during typhoon season, characterized in that, include: Acquire images of waves crossing during typhoon seasons and waves climbing during normal seasons, and classify the collected images of waves crossing during typhoon seasons and waves climbing during normal seasons. The image of the wave rise during the normal period is processed to obtain a semantic segmentation binary image of the wave region and the first background region; The images of typhoon-era waves are processed to obtain semantic segmentation binary maps of the wave-over region and the second background region, and the main body of the dam in the second background region is identified. The main body of the dam is divided into multiple sections, and the dam height and length of each section are measured to obtain section information. RTK data of the main body of the dam is collected, and interpolation is performed on the RTK data to obtain the pixel points of different survey lines. The dam height corresponding to each section and the pixel corresponding to the elevation of the dam height are obtained to form a mapping relationship between dam height and pixels. The first most convex point of the wave region and the second most convex point of the wave overshoot region are extracted. The calibration information of one section corresponding to the first or second most convex point is selected from the multiple section information, and the wave climb point or the highest point of the wave overshoot is obtained based on the calibration information.

2. The typhoon wave run-up identification method according to claim 1, characterized in that, The collected images of wave overtopping during typhoon seasons and wave rise during normal seasons were categorized, including: A classification model was built using the improved EfficientNetV2 algorithm; The collected images of typhoon-period wave overpasses and normal-period wave rises are input into the classification model for classification, and the classified images of typhoon-period wave overpasses and normal-period wave rises are output.

3. The method for identifying wave run-up during typhoon season according to claim 1, characterized in that, Also includes: The identified dam body is marked with a rectangular frame.

4. The typhoon wave rise identification method according to claim 3, characterized in that, Also includes: Obtain the overlapping area between the wave-crossing area and the main body of the dam, and calculate the area ratio of the overlapping area on the main body of the dam; When the area ratio exceeds a set threshold, the dwell time of the current overtopping area on the main body of the embankment is obtained. If the dwell time exceeds the set time, it is determined to be a severe overtopping.

5. The method for identifying wave run-up during typhoon season according to claim 1, characterized in that, The improved PSPNet algorithm was used to process images of wave rise during the normal period, the improved UET algorithm was used to process images during the typhoon period, and the improved YOLOv3 algorithm was used to identify the main body of the dam in the second background region.

6. A wave run-up identification device during typhoon season, characterized in that, include: Image collection module: used to acquire images of waves crossing during typhoon season and waves climbing during normal season, and to classify the collected images of waves crossing during typhoon season and waves climbing during normal season; The first processing module is used to process the wave rise image during the normal period to obtain a semantic segmentation binary image of the wave region and the first background region. The second processing module is used to process the images of typhoon-era waves to obtain semantic segmentation binary images of the wave-overlapping area and the second background area, and to identify the main body of the dam in the second background area. High-point acquisition module: This module divides the main body of the dam into multiple sections, measures the dam height and length of each section to obtain section information; collects RTK data of the main body of the dam, and performs interpolation and sorting based on the RTK data to obtain the pixel points of different survey lines; obtains the dam height corresponding to each section and the pixel corresponding to the elevation of that dam height, forming a mapping relationship between dam height and pixels; extracts the first most convex point of the wave region and the second most convex point of the wave-overtaking region; selects the calibration information of one section information corresponding to the first or second most convex point from multiple section information, and obtains the wave climb point or the highest point of the wave overtaking based on the calibration information.

7. An electronic device, characterized in that, include: Memory containing executable program code; A processor coupled to the memory; The processor calls the executable program code stored in the memory to execute the typhoon wave rise identification method according to any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, wherein the computer program causes a computer to execute the typhoon wave rise identification method according to any one of claims 1 to 5.

Images