Overwave detection method and system

Abstract

This application discloses a wave overrun detection method and system, belonging to the field of coastal engineering disaster detection. The method includes: collecting raw point cloud data and video data at preset time intervals; calculating the spatial angle of each beam emission direction; constructing a beam angle grid with preset angle intervals; and calculating the maximum detection distance of the scanning device within each grid to obtain the maximum distance mapping relationship of the beam direction. Based on the maximum distance mapping relationship of the beam direction, a point cloud direct processing method without spatiotemporal interpolation is used to complete the maximum beam distance detection, point cloud static and dynamic classification, shoreline detection, and wave run-up parameter calculation. The wave run-up parameters and wave overrun detection results are output, and corresponding synchronous video frames are automatically retrieved and output based on the maximum run-up height or key nodes where wave overrun occurs. This application has strong environmental adaptability, can perform continuous detection throughout the day, has high wave run-up measurement accuracy, and features flexible and easy-to-maintain hardware deployment, making it valuable for engineering applications and promotion.

Description

Technical Field

[0001] This invention relates to the field of coastal engineering disaster detection technology, and in particular to a method and system for detecting overtopping waves. Background Technology

[0002] Wave rise and overtopping are major contributing factors to marine disasters such as storm surges and giant waves, easily causing inland flooding and damage to coastal protection structures, posing a serious threat to the safety of coastal personnel and property. Real-time and accurate wave rise and overtopping detection is a core component in building a coastal disaster early warning system and formulating disaster prevention and mitigation measures, providing crucial information for local management departments to issue evacuation warnings and carry out emergency response.

[0003] Currently, wave overtopping detection mainly relies on optical equipment such as surveillance cameras. However, although such optical equipment has the advantages of low deployment cost and simple operation, as passive sensing devices, it is limited by lighting conditions and cannot work normally under harsh conditions such as night, heavy rain, heavy snow, and steam fog, resulting in poor detection continuity. Its water level estimation relies on known beach topographic profiles, and in coastal areas with dynamic topographic changes, frequent topographic measurements are required, making it difficult to achieve long-term continuous detection. More importantly, it can only provide qualitative visual information on coastal water conditions and cannot obtain quantitative data such as wave rise height and overtopping intensity, making it difficult to meet the needs of refined coastal disaster management.

[0004] In recent years, lidar technology has been explored for wave crossing detection due to its active detection and high-precision ranging capabilities. However, existing lidar-based detection technologies still have significant shortcomings: point cloud data acquired by 3D lidar is susceptible to noise from rain, snow, fog, and coastal protection structures, and existing filtering methods are complex and have poor adaptability; traditional lidar data processing relies on spatiotemporal interpolation algorithms, which are computationally expensive and inefficient, failing to meet the timeliness requirements of real-time detection; and existing technologies are mostly developed for ideal laboratory conditions or short-term field tests, failing to form an integrated real-time detection system adapted to complex coastal environments and capable of long-term stable operation, making it difficult to directly apply to practical engineering projects.

[0005] Therefore, how to overcome interference from severe weather and complex terrain, achieve efficient and low-cost lidar point cloud data processing, and obtain continuous, real-time, and quantitative detection data on wave rise and overtopping remains a problem that urgently needs to be solved in this field. Summary of the Invention

[0006] The purpose of this application is to overcome the shortcomings of the existing technology and provide a wave overrun detection method and system that can continuously and quantitatively detect the wave rise height and wave overrun phenomenon in the coastal zone under complex coastal topography and severe weather conditions, providing data support for coastal disaster early warning, coastal protection engineering operation and maintenance and coastal management.

[0007] Firstly, this application provides a wave-crossing detection method, comprising the following steps: Raw point cloud data is collected at preset time intervals. For each batch of raw point cloud data, the spatial angle of each beam emission direction is calculated. A beam angle grid is constructed at preset angle intervals, and the maximum detection distance of the scanning device within each grid is calculated to obtain the maximum distance mapping relationship of the beam direction. Based on the maximum distance mapping relationship of the beam direction, the original point cloud data is converted to the local coastal coordinate system and divided into grids. The beach bed elevation of each grid cell is determined. The point cloud data under the local coastal coordinate system is classified into static points and dynamic points. The water depth value of the dynamic points relative to the beach bed elevation is calculated. Dynamic points are sliced ​​in time at preset time intervals, and shoreline candidate points are screened and spatial neighborhood constraints are applied to the dynamic points after time-series slicing to obtain initial shoreline positions. Adaptive threshold correction is applied to the initial shoreline positions to obtain time-series shoreline elevation data. Wave run-up parameters are calculated based on the time-series shoreline elevation data, and the wave run-up parameters and wave overrun detection results are output.

[0008] Optionally, determining the beachbed elevation of each grid cell includes: Based on the maximum distance mapping relationship of the beam direction, according to the beam emission direction corresponding to each grid cell, its maximum detection distance in the maximum distance mapping relationship of the beam direction is indexed; the elevation of the spatial point corresponding to the maximum detection distance is regarded as the candidate static target elevation of the grid cell; after traversing all grid cells, spatial filtering is performed on the candidate static target elevation to remove outliers and obtain the static elevation distribution of each grid cell; The elevation values ​​of all transformed point cloud data within each grid cell are extracted, and their statistical quantiles are calculated. The statistical quantiles are then compared and fused with the static elevation distribution to obtain the fused reference elevation. After neighborhood smoothing to eliminate local abnormal fluctuations, the beach bed elevation of each grid cell is obtained.

[0009] Optionally, the point cloud data in the local coastal coordinate system is classified into static points and dynamic points, and the water depth values ​​of the dynamic points relative to the beachbed elevation are calculated, including: Based on the static elevation distribution, the initial static elevation interval is formed by extending upward and downward by a first preset tolerance value. With the beach bed elevation as a reference, the lower boundary of the initial static elevation interval is constrained by a second preset tolerance value to obtain the reasonable range of the static target for each grid cell. Points whose elevation values ​​are within the reasonable range of the static target are determined as static points; the remaining points are determined as dynamic points.

[0010] Optionally, the step of performing time-series slicing on dynamic points at preset time intervals, and then performing shoreline candidate point screening and spatial neighborhood constraint processing on the time-series sliced ​​dynamic points to obtain initial shoreline positions, followed by adaptive threshold correction processing on the initial shoreline positions to obtain time-series shoreline elevation data, includes: The dynamic points are processed by time-series slicing at preset time intervals to obtain multiple time-series slice dynamic point sets; The time-series slice dynamic point set is subjected to water depth threshold screening to obtain each initial shoreline candidate point set; Spatial neighborhood constraints are applied to each initial shoreline candidate point set. From the candidate points that satisfy the spatial neighborhood constraints, the point with the smallest coordinate value along the transverse shoreline direction and that satisfies the constraints on the number of dynamic points in the neighborhood and the consistency of elevation is selected to obtain the initial shoreline position and its three-dimensional coordinates for each time slice. Determine whether the distance between each initial shoreline location and the coastal protection structure is less than a preset distance threshold; If the distance is less than the preset distance threshold, the water depth threshold range is adaptively adjusted, and dynamic points are re-screened in the area in front of the coastal protection structure to obtain the corrected shoreline candidate point set and the corrected shoreline position. If the distance is greater than or equal to the preset distance threshold, the current shoreline position is taken as the final shoreline position. The final shoreline positions of all time-series slices are integrated in chronological order to obtain time-series shoreline elevation data covering the entire collection period.

[0011] Optionally, spatial neighborhood constraints are applied to each initial shoreline candidate point set. From the candidate points that satisfy the spatial neighborhood constraints, the point with the smallest coordinate value along the transverse shoreline direction and that satisfies the constraints on the number of dynamic points in the neighborhood and the consistency of elevation is selected to obtain the initial shoreline position and its three-dimensional coordinates for each time-series slice, including: Taking any candidate point in each initial shoreline candidate point set as the center, a three-dimensional spatial neighborhood is defined in the local coastal coordinate system with a first radius as a spatial neighborhood constraint. The number of dynamic points within the spatial neighborhood constraint is counted. When the number is greater than or equal to a preset threshold for the number of dynamic points in the neighborhood, the candidate point enters the next level of constraint judgment. Calculate the elevation standard deviation of all dynamic points within the spatial neighborhood constraint. When the elevation standard deviation is less than or equal to a preset elevation consistency threshold, the candidate point satisfies all the constraint conditions. From all candidate points that satisfy all constraints, select the candidate point with the smallest coordinate value along the transverse bank direction as the initial shoreline position of the time slice, and obtain its three-dimensional coordinates.

[0012] Optionally, it also includes: calculating wave run-up parameters based on the time-series shoreline elevation data, wherein the wave run-up parameters include maximum run-up height, characteristic run-up height and average shoreline elevation, and when the maximum run-up height is greater than or equal to the top elevation of the coastal protection structure, it is determined to be a wave overrun, and the time and duration of the overrun are recorded. Output wave climb parameters and wave overtopping detection results, and automatically retrieve and output corresponding synchronous video frames based on the maximum climb height or key nodes where wave overtopping occurs.

[0013] Optionally, wave run-up parameters are calculated based on the time-series shoreline elevation data, including: The local maximum value of shoreline elevation is extracted using a fixed time window as the single-wave climb height, and the maximum value among all single-wave climb heights is the maximum climb height within the fixed time window. Sort all single-wave climbing heights from largest to smallest, and take the corresponding values ​​with a cumulative probability greater than 2% as the feature climbing heights; Calculate the arithmetic mean of all time-series shoreline elevation data within the fixed time window, and use it as the average shoreline elevation.

[0014] Secondly, this application also provides a wave rise and overtopping detection system, comprising: The hardware deployment module is used to acquire raw point cloud data and video data of the detection area; The data synchronization acquisition module performs spatiotemporal synchronization acquisition of the original point cloud data and video data to obtain spatiotemporal synchronized point cloud data and spatiotemporal synchronized video data, and has an anomaly detection function. The lidar point cloud processing module adopts a direct point cloud processing method without spatiotemporal interpolation. It sequentially performs maximum distance detection, static and dynamic point cloud classification, and shoreline position detection on the spatiotemporal synchronized point cloud data to obtain the time-series shoreline elevation, and calculates wave run-up parameters and wave overtopping determination results accordingly. The detection result output module outputs the wave climb parameters and wave overshoot determination results. It also visualizes the results by linking and calling the spatiotemporally synchronized video data based on key nodes, and stores the detection results locally and uploads them remotely. All modules work collaboratively to achieve integrated real-time detection from data acquisition and processing to result output.

[0015] Optionally, the lidar point cloud processing module adopts a direct point cloud data processing method that does not require spatiotemporal interpolation. Through three core processes—maximum distance detection of beam direction, static and dynamic classification of point cloud data, and shoreline position detection and wave detection parameter calculation—it performs noise filtering, shoreline identification, and rapid calculation of wave rise and overtopping parameters.

[0016] Optionally, the hardware deployment module includes: Coastal inspection pole, which serves as a supporting carrier; The scanning device is installed at a first preset height on the coastal detection pole. It is used to receive laser echo signals from the coastal detection area, perform active scanning and ranging processing, and output raw three-dimensional point cloud data. The monitoring equipment is installed at a second preset height on the coastal detection pole, and its field of view is completely overlapped with that of the scanning equipment after spatial calibration. It is used to receive visible light signals, perform lens distortion correction and video encoding processing, and output synchronous video stream data. The data processing terminal device is deployed at a third preset height on the coastal detection pole. It is used to receive the original three-dimensional point cloud data and synchronous video stream data, perform real-time data processing, parameter calculation and result storage, and remotely upload the final detection results to the coastal disaster detection cloud platform via wireless or wired means.

[0017] This application provides a wave overtopping detection method and system that abandons the spatiotemporal interpolation operation of traditional data processing and adopts a direct point cloud data processing method, which greatly simplifies the processing flow, reduces computing costs, and the average processing time of a single data acquisition is much shorter than the acquisition time. It can realize real-time processing and result output of detection data, meet the timeliness requirements of coastal disaster early warning, and has strong environmental adaptability, efficient data processing, and accurate detection results.

[0018] To make the above-mentioned features and advantages of the invention more apparent and understandable, specific embodiments are described below, and detailed descriptions are provided in conjunction with the accompanying drawings. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments or related technologies of this application, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a schematic diagram of the structure of a wave detection system provided in one embodiment of this application.

[0021] Figure 2 This is a schematic diagram of the hardware deployment module in the wave run-up and overtopping detection system of this application.

[0022] Figure 3 for Figure 2 A schematic diagram of area A in the hardware deployment module shown.

[0023] Figure 4 for Figure 2 A schematic diagram of area B in the hardware deployment module shown.

[0024] Figure 5 This is a flowchart of a wave detection method provided in another embodiment of this application.

[0025] Figure 6 This is a schematic diagram illustrating the principle of mapping the maximum detection distance to the beam emission direction.

[0026] Figure 7 A schematic diagram of obtaining the local coastal coordinate system and cross-sectional point cloud.

[0027] Figure 8 This is a schematic diagram illustrating the principles of point cloud classification and water depth calculation.

[0028] Figure 9 This is a flowchart of step S3 in a wave detection method provided in one embodiment of this application. Detailed Implementation

[0029] To make the objectives and technical solutions of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the described embodiments of this application without creative effort are within the scope of protection of this application.

[0030] In one embodiment, see Figure 1 This application provides a wave overrun detection system, which may include: a hardware deployment module 1, used to acquire raw point cloud data and video data of the detection area; a data synchronization acquisition module 2, which performs spatiotemporal synchronous acquisition of the raw point cloud data and video data to obtain spatiotemporal synchronous point cloud data and spatiotemporal synchronous video data, and has an anomaly detection function; a lidar point cloud processing module 3, which adopts a direct point cloud processing method without spatiotemporal interpolation, sequentially performs maximum distance detection, static and dynamic point cloud classification, and shoreline position detection on the spatiotemporal synchronous point cloud data to obtain time-series shoreline elevation, and calculates wave rise parameters and wave overrun determination results accordingly; and a detection result output module 4, which outputs the wave rise parameters and wave overrun determination results, links the spatiotemporal synchronous video data according to key nodes for visualization display, and performs local storage and remote uploading of the detection results.

[0031] For example, please refer to Figure 2The hardware deployment module 1 includes: a coastal detection pole 14, which serves as a supporting carrier; a scanning device 11, installed at a first preset height on the coastal detection pole 14, used to receive laser echo signals from the coastal detection area, perform active scanning and ranging processing, and output raw three-dimensional point cloud data; a monitoring device 12, installed at a second preset height on the coastal detection pole 14, whose field of view is completely overlapped with that of the scanning device 11 after spatial calibration, used to receive visible light signals, perform lens distortion correction and video encoding processing, and output synchronous video stream data; and a data processing terminal device 13, deployed at a third preset height on the coastal detection pole 14, used to receive the raw three-dimensional point cloud data and synchronous video stream data, perform real-time data processing, parameter calculation and result storage, and remotely upload the final detection results to the coastal disaster detection cloud platform via wireless or wired means. The third preset height is smaller than the first and second preset heights.

[0032] Figure 3 for Figure 2 The enlarged view of area A in hardware deployment module 1 shows that scanning device 11 and monitoring device 12 are mounted together on the coastal detection pole 14. Their installation height and viewing angle are matched to ensure that their detection field of view completely covers the target beach and coastal protection structure.

[0033] As an example, the spatial position and viewing angle calibration of the scanning device 11 and the monitoring device 12 can be completed by measuring the ground control points in the detection area through the real-time dynamic global navigation satellite system (RTK-GNSS); the accuracy calibration of the point cloud data of the scanning device 11 and the lens distortion correction and internal parameter calibration of the monitoring device 12 can be completed by using a standard target.

[0034] For example, the scanning device 11 can be a three-dimensional lidar scanner. The three-dimensional lidar scanner is a cone-shaped field-of-view scanning device with a high point cloud acquisition rate, which can accurately capture subtle changes in coastal topography and dynamic changes in the water surface, and the scanning field of view covers the entire beach and coastal protection structure area of ​​the detection area.

[0035] For example, the monitoring device 12 can be a high-definition surveillance camera. The high-definition surveillance camera has lens distortion correction capabilities, and its recording resolution and frame rate are adapted to the needs of coastal detection, enabling it to clearly capture visual information about the coastal water conditions in the detection area.

[0036] For example, please refer to Figure 4The data processing terminal device 13 includes: a computer 131 for real-time processing of the collected data; a hard disk drive 132 connected to the computer 131 for local large-capacity storage of the processed detection data and backup of the original data; and a protective case 133 for housing the computer 131 and the hard disk drive 132, providing a protective environment against salt spray and humidity for the internal equipment.

[0037] As an example, the protective box 133 is a sealed protective box.

[0038] As an example, the data processing terminal device 13 also includes a power supply voltage regulator and protective accessories (not shown), which are deployed inside the protective box 133 and utilize the existing power supply system on the coast to achieve continuous power supply.

[0039] For example, the data processing terminal device 13 is embedded, with high-speed data processing and large-capacity storage capabilities, and can complete real-time data processing and local storage.

[0040] As an example, the equipment in hardware deployment module 1 is all modularly designed and can be deployed using existing coastal inspection poles and power supply systems, making it easy to assemble and maintain on site.

[0041] As an example, all equipment is adapted to the high salt spray and high humidity environment of the coast.

[0042] As an example, the data synchronization acquisition module 2 integrates the development interface of the scanning device 11 and the machine vision open source library to develop a dedicated data synchronization acquisition program, realizing the time and space dual synchronous acquisition of point cloud data from the scanning device 11 and video data from the monitoring device 12. It adopts a timed segmented acquisition strategy, starting data acquisition at fixed time intervals. The duration of each acquisition is set according to the dynamic detection requirements of coastal waves. The acquired data is transmitted to the data processing terminal for local storage in real time to avoid data loss. The acquisition program has an anomaly detection function, which can identify abnormal situations such as power failure, missing point cloud data, and camera obstruction in real time, and mark and warn locally.

[0043] For example, the data synchronization acquisition module 2 starts the data synchronization acquisition program and starts data acquisition at a preset time interval. The duration of a single acquisition is set to 10 minutes. The program automatically realizes the time and space dual synchronous acquisition of LiDAR point cloud data and camera video data. The acquired raw data is transmitted to the data processing terminal device 13 for local storage in real time. At the same time, the acquisition program monitors the device's operating status in real time. If an abnormality is detected, it is immediately marked locally and an early warning is issued.

[0044] As an example, the lidar point cloud processing module 3 adopts a direct processing method for point cloud data without spatiotemporal interpolation. Through three core processes—maximum distance detection of beam direction, static and dynamic classification of point cloud data, and shoreline position detection and wave detection parameter calculation—it achieves noise filtering, shoreline identification, and rapid calculation of wave rise and overtopping parameters.

[0045] As an example, the maximum beam direction distance detection process involves calculating the spatial angle of the beam emission direction and gridding it, generating a mapping relationship of the maximum beam direction distance from the scanning device to the detection target within each grid. The static and dynamic point cloud data classification process involves converting the point cloud data to the local coastal coordinate system and gridding it, determining the beachbed elevation based on the maximum beam direction distance mapping relationship, dividing the point cloud data into static and dynamic points, and calculating the water depth value of the dynamic points relative to the beachbed elevation. The shoreline location detection and wave detection parameter calculation process involves selecting dynamic points with water depth values ​​within a preset threshold range as shoreline candidate points based on time-series slices, selecting the landward candidate point as the initial shoreline position through spatial neighborhood constraints, adaptively adjusting the water depth threshold to complete shoreline correction when approaching coastal protection structures, and calculating the maximum climb height based on time-series shoreline elevation data. Characteristic climbing height Average shoreline elevation When the maximum climbing height When the wave reaches the top elevation of the coastal protection structure, it is determined to be a wave overtopping event, and the time of occurrence and duration are recorded.

[0046] As an example, the detection result output module 4 can automatically retrieve and output synchronized video frames from the camera based on key nodes such as wave rise extremes and wave overtopping occurrences detected by the lidar. The processed detection data is uploaded to the coastal disaster detection cloud platform in real time, supporting remote access, querying, and visualization. At the same time, the embedded data processing terminal locally retains historical data backups, and the backup cycle can be set according to requirements.

[0047] As an example, the system in this application is compatible with various coastal types such as gravel beaches and mixed gravel beaches, and can work stably in complex environments such as heavy rain, heavy snow, steam fog, and coastal protection structures such as breakwaters and wave-dissipating blocks.

[0048] The aforementioned wave overtopping detection system is flexible in deployment and highly adaptable to engineering applications. It can be deployed using existing coastal detection poles and power supply systems, eliminating the need for large-scale infrastructure construction. Furthermore, its core equipment features a modular design with a high degree of standardization, facilitating on-site assembly and maintenance, and making it suitable for various types of coastal detection areas. It boasts excellent data fusion capabilities and comprehensive detection information, achieving synchronous linkage between quantitative detection data from lidar and qualitative visual data from cameras. Corresponding video information can be quickly retrieved through key detection nodes, providing comprehensive detection data for tracing and analyzing coastal disaster events, thus enhancing the practicality of the detection results. This system can operate stably under different coastal types and complex weather and terrain conditions, achieving real-time, accurate, and quantitative detection of wave rise and overtopping. The system exhibits high data processing efficiency and reliable detection results, demonstrating excellent engineering applicability and promotional value.

[0049] In another embodiment, please refer to Figure 5 This application provides a wave detection method, which may include the following steps: steps S1 to S6.

[0050] Step S1: Collect raw point cloud data at preset time intervals. For each batch of raw point cloud data, calculate the spatial angle of each beam emission direction. Construct a beam angle grid with preset angle intervals and calculate the maximum detection distance of the scanning device within each grid to obtain the maximum distance mapping relationship of the beam direction.

[0051] Step S2: Based on the maximum distance mapping relationship of the beam direction, the original point cloud data is converted to the local coastal coordinate system and divided into grids. The beach bed elevation of each grid cell is determined. The point cloud data under the local coastal coordinate system is classified into static points and dynamic points. The water depth value of the dynamic points relative to the beach bed elevation is calculated.

[0052] Step S3: Perform time-series slicing on the dynamic points at preset time intervals, and perform shoreline candidate point screening and spatial neighborhood constraint processing on the dynamic points after time-series slicing to obtain the initial shoreline position. Perform adaptive threshold correction processing on the initial shoreline position to obtain time-series shoreline elevation data.

[0053] Step S4: Calculate wave run-up parameters based on the time-series shoreline elevation data. The wave run-up parameters include maximum run-up height, characteristic run-up height, and average shoreline elevation. When the maximum run-up height is greater than or equal to the top elevation of the coastal protection structure, it is determined to be a wave overrun. Record the time of the overrun and its duration.

[0054] Step S5: Output wave climb parameters and wave overshoot detection results, and automatically retrieve and output the corresponding synchronous video frames based on the maximum climb height or the key node where the wave overshoot occurs.

[0055] In the wave overrun detection method of this application, invalid noise points are effectively filtered out and the quality of the original point cloud is improved by constructing a beam angle grid and mapping it to the maximum detection distance. The point cloud is transformed to the local coastal coordinate system and static and dynamic points are classified. Combined with water depth calculation and time-series slice processing, dynamic tracking of the wave rise process is realized. The beach bed elevation and water depth are calculated based on the original point cloud data. Based on the time-series slices, shoreline candidate points are screened and adaptive threshold correction is performed on the dynamic points. This simplifies the point cloud data processing process, abandons the traditional complex spatiotemporal interpolation operation, reduces the computational cost, improves data processing efficiency, and meets the real-time requirements of disaster detection.

[0056] As an example, before step S1, the following steps are also included: modular deployment of the scanning device 11, the monitoring device 12, and the data processing terminal device 13 on the coast.

[0057] Specifically, based on the scope and topographic features of the coastal detection area, a suitable coastal detection pole 14 is selected. The scanning device 11 and monitoring device 12 are co-mounted on the coastal detection pole 14. The precise coordinates of at least three fixed ground control points within the detection area are measured using a real-time dynamic global navigation satellite system. These control points are used to calibrate the spatial position and viewing angle of the scanning device 11 and monitoring device 12, achieving spatial synchronization of their data. Simultaneously, a standard target is used to calibrate the accuracy of the point cloud data from the scanning device 11 and to correct lens distortion and calibrate the intrinsic parameters of the monitoring device 12, ensuring complete overlap of their detection fields of view. The data processing terminal device 13 is then sealed and protected, and its hardware is connected. Based on the coastal topography, wave characteristics, and protective structure parameters of the detection area, the core operating parameters of the system are preset, resulting in a spatially synchronized integrated detection system.

[0058] As an example, the scanning device 11 may be a three-dimensional LiDAR scanner.

[0059] As an example, the monitoring device 12 includes, but is not limited to, a high-definition surveillance camera.

[0060] As an example, the data processing terminal device 13 is an embedded data processing terminal device.

[0061] As an example, the accuracy calibration of the point cloud data of the scanning device 11 using a standard target includes: placing a standard target of known geometry, such as a sphere or a planar target, in the detection area; the scanning device 11 acquiring point cloud data; extracting target feature points; comparing the target with the actual target parameters to calculate the error; and applying calibration parameters to correct the accuracy of the point cloud data.

[0062] As an example, the lens distortion correction and intrinsic parameter calibration of the monitoring device 12 are completed, including: the monitoring device 12 captures a standard target image, extracts target feature points using an image processing algorithm, calculates lens distortion parameters and camera intrinsic parameter matrix, and realizes image distortion correction and geometric calibration.

[0063] Furthermore, based on the coastal topography, wave characteristics, and protective structure parameters of the detection area, core parameters such as time intervals and angle intervals are preset in the data processing terminal.

[0064] It is important to know that all equipment is adapted to the high salt spray and high humidity environment of the coastal zone and has structural stability that is resistant to wind and waves.

[0065] In step S1, please refer to Figure 5 In step S1, raw point cloud data is collected at preset time intervals. For the raw point cloud data collected in a single instance, the spatial angle of each beam emission direction is calculated. A beam angle grid is constructed at preset angle intervals, and the maximum detection distance of the scanning device within each grid is calculated to obtain the maximum distance mapping relationship of the beam direction.

[0066] Specifically, the scanning device 11 and the monitoring device 12 are synchronously acquired in both time and space at preset time intervals to obtain raw point cloud data and video data. At the same time, real-time anomaly detection is performed on the device's operating status. If an anomaly is detected, it is immediately marked locally and an early warning is issued, obtaining synchronized detection data and device status information.

[0067] As an example, the raw point cloud data and video data are spatiotemporally synchronized. Spatiotemporal synchronization can be achieved by marking each frame of point cloud data and each frame of video data with a uniform timestamp.

[0068] As an example, the preset time interval can be determined according to wave detection requirements. For instance, it can be set to start data acquisition every 20 minutes, with each acquisition lasting 10 minutes. In this way, the data acquired each time constitutes an independent processing unit. It is understood that the preset time interval and acquisition duration are not limited to these and can be flexibly configured according to the wave and tide characteristics of the detection area and storage resources.

[0069] As an example, device anomalies include, but are not limited to, abnormal situations such as device power failure, missing point cloud data, and camera obstruction.

[0070] Furthermore, for the raw point cloud data acquired in a single acquisition, the spatial angle of the beam emission direction of each point is calculated using the pitch and azimuth information of each point in the original coordinate system of the scanning device.

[0071] For example, based on raw point cloud data from a single acquisition, the coordinates of a certain point in the original coordinate system of the scanning device are: The horizontal azimuth angle of the beam emission direction is The vertical pitch angle is Then the point is obtained. The only corresponding set of spatial angles is .

[0072] Furthermore, all beam emission directions are discretized at preset angular intervals to construct a beam angle grid. For all raw point cloud data acquired during the acquisition period, each grid cell is traversed, and the point with the largest distance from the laser emission origin is selected. This distance value is recorded as the maximum detection distance of the corresponding grid cell. After traversing all grid cells, a complete maximum distance mapping relationship of the beam direction is obtained.

[0073] As an example, the maximum distance mapping relationship of the beam direction refers to the correspondence established between the beam emission angle grid and the farthest effective echo distance detected in that angle direction. The maximum detection distance in the maximum distance mapping relationship of the beam direction reflects the actual detection distance of the beam when it first encounters a solid target (such as a beach surface, shoreline topography, protective structure, etc.) in the corresponding emission direction, providing a key reference benchmark for subsequent point cloud classification and a basis for static target detection.

[0074] As an example, the preset angle interval can be set to a uniform grid of 0.04° × 0.04°, which is the grid interval for the horizontal azimuth angle. Grid spacing equal to vertical pitch angle It equals 0.04°. Other intervals can also be selected based on the angular resolution of the acquired point cloud. Each grid cell corresponds to a narrow range of beam emission directions.

[0075] Figure 6 This is a schematic diagram illustrating the principle of mapping the maximum detection distance of the beam emission direction. The left side shows a scanning device mounted on a coastal detection pole, which scans the field of view vertically. Multiple beams of light at different spatial angles are emitted into the detection area (shown as blue rays in the figure). Their scanning boundaries (shown as solid black lines) completely cover the target beachbed and coastal protection structures (shown as gray retaining walls in the figure). The blue dashed lines represent the maximum detection distance of static solid targets in different beam directions. During a single data acquisition period, for any beam with a fixed emission direction, the static solid targets (beachbed, coastal protection structures) along its detection path are always behind the dynamic targets (water surface, waves, foam). Therefore, the maximum detection distance in that beam direction necessarily corresponds to the actual distance of the static solid targets. Based on this characteristic, this application does not require prior acquisition of static terrain data of the detection area, nor does it require complex spatiotemporal interpolation calculations; the baseline distance information of static targets can be extracted solely from the point cloud data acquired in a single session.

[0076] The above maximum distance mapping relationship of beam direction does not depend on any spatiotemporal interpolation operation. It does not use time accumulation interpolation of the original sparse point cloud to generate a dense depth map, nor does it require interpolation to fill spatial holes. Instead, it is obtained directly based on the spatial angle statistics of the original point cloud data, which preserves the real detection information to the greatest extent and reduces the computational complexity.

[0077] In step S2, please refer to Figure 5 In step S2, the original point cloud data is converted to the local coastal coordinate system and divided into grids according to the maximum distance mapping relationship of the beam direction. The beach bed elevation of each grid cell is determined, and the point cloud data under the local coastal coordinate system is classified into static points and dynamic points. The water depth value of the dynamic points relative to the beach bed elevation is calculated.

[0078] Specifically, a local coastal coordinate system is established with a fixed reference point on the detected shoreline as the origin. The x-axis points positive along the transverse direction of the coast and towards the sea; the y-axis points positive along the longitudinal direction of the coastline and towards the main direction of the coastal current; and the z-axis points vertically upward and is referenced to the local mean sea level, forming a right-handed rectangular coordinate system. All raw point cloud data collected in a single session are transformed from the polar coordinate system of the scanning device to the local coastal coordinate system.

[0079] Furthermore, transverse and longitudinal shore grids are constructed at preset intervals in the local coastal coordinate system, with each grid unit covering a certain range of transverse and longitudinal shore areas.

[0080] As an example, the preset interval can be set to 0.05m.

[0081] Furthermore, based on the maximum distance mapping relationship of the beam direction, according to the beam emission direction corresponding to each grid cell, its maximum detection distance in the maximum distance mapping relationship of the beam direction is indexed; the elevation of the spatial point corresponding to the maximum detection distance is regarded as the candidate static target elevation of the grid cell. After traversing all grid cells, spatial filtering is performed on the candidate static target elevation to remove outliers caused by wave splash, sea foam, etc., to obtain the static elevation distribution of each grid cell.

[0082] Furthermore, the elevation values ​​of all transformed point cloud data within each grid cell are extracted, arranged in ascending order, and their statistical quantiles are calculated. The statistical quantiles are then compared and fused with the static elevation distribution to obtain the fused reference elevation. After neighborhood smoothing to eliminate local abnormal fluctuations, the beachbed elevation of each grid cell is obtained.

[0083] As an example, the statistical quantile can be the p-th quantile. The quantile ratio p is selected based on the beach topography and the signal-to-noise ratio of the point cloud. For example, p=5%, which means taking the elevation values ​​in the top 5% after sorting. This value can effectively represent the beach surface or protective structure surface touched by the laser after penetrating waves and foam, while avoiding excessively low abnormal elevations introduced by a few noise points, splashing water droplets, etc. It is understood that the value of the quantile ratio p is not limited to 5%, and other appropriate ratios such as 3% or 10% can also be selected according to the actual wave conditions and the required level of conservatism.

[0084] As an example, the specific method of comparison and fusion is as follows: if the difference between the statistical quantile and the static elevation within a certain grid cell is less than or equal to a preset deviation threshold, it indicates that the two are consistent, and the weighted average of the two or the statistical quantile is taken directly as the fused reference elevation of the grid cell; if the difference between the two is greater than the preset deviation threshold, it indicates that the grid cell may be affected by occlusion, dense dynamic targets, or edge effects of structures, etc., in which case the statistical quantile is given priority, supplemented by the static elevation for rationality verification, and the fused reference elevation of the grid cell is determined after correction.

[0085] As an example, the preset deviation value can be 0.15m.

[0086] As an example, the neighborhood smoothing process can be implemented as follows: a preset neighborhood window is constructed centered on each grid cell, and the average or weighted median of all merged reference elevations within the preset neighborhood window is calculated. For grid cells near coastal protection, such as breakwaters and wave-dissipating blocks, due to their drastic elevation changes and reasonable characteristics of abrupt topographic changes, their original merged reference elevations can be identified and retained before smoothing. The smoothing operation is performed only on grids in natural beach areas, thereby removing noise while maintaining the true topographic features of the protective structure's edges.

[0087] As an example, the preset neighborhood window can be set to a 3×3 or 5×5 grid.

[0088] Furthermore, based on the static elevation distribution, the initial static elevation interval is formed by extending upward and downward by a first preset tolerance value. With the beachbed elevation as a reference, the lower boundary of the initial static elevation interval is constrained by a second preset tolerance value to obtain the reasonable range of the static target of the grid cell. Points whose elevation values ​​are within the reasonable range of the static target are determined as static points; the remaining points are determined as dynamic points.

[0089] As an example, the static point can be a fixed target such as a beach surface or a coastal protection structure.

[0090] As an example, the dynamic point can be an instantaneous target such as the water surface, waves, foam generated by wave breaking, and splashing water droplets.

[0091] Furthermore, for each point classified as a dynamic point, the water depth value relative to the beachbed elevation is calculated. The water depth value is equal to the elevation value of the dynamic point minus the beachbed elevation of the corresponding grid cell. A positive water depth value indicates that the dynamic point is above the beachbed, corresponding to the instantaneous water surface position during wave rise; a negative water depth value indicates that the dynamic point is below the beachbed elevation, and can be marked or filtered out in subsequent processing.

[0092] Figure 7 This diagram illustrates the acquisition of point cloud data along the local coastal coordinate system and cross-section. Orange dots represent the raw point cloud data acquired by the scanning device along the target cross-section. The scanning device emits multiple beams at different spatial angles into the detection area, acquiring continuously distributed raw point cloud data along the xz plane of the target cross-section. Its scanning range completely covers the coastal protection structure and nearshore beach surface area. All orange dots show the dynamic change trajectory of the shoreline position during wave rise.

[0093] Figure 8 This diagram illustrates the principle of point cloud classification and water depth calculation. Orange and green dots represent static and dynamic points, respectively, while blue dots represent the instantaneous shoreline position obtained based on the dynamic point water depth detection. It can be seen that for any beam emitted along a fixed direction, static solid targets along its detection path are always located behind dynamic water targets. Therefore, in the same beam direction, the point with the lowest elevation is the surface point of the static solid target, and its elevation value is the beachbed elevation at that location. All points above this point are dynamic water points, and the difference between their elevations and the corresponding beachbed elevations is the instantaneous water depth value at that point. The instantaneous shoreline obtained based on the dynamic point water depth detection is located at the interface between the static beachbed and the dynamic water body, corresponding to the instantaneous waterline at wave rise.

[0094] The above steps utilize the maximum distance mapping relationship of the beam direction to directly perform dynamic and static binary classification on the original point cloud data, assigning dynamic point depth values. This completely abandons the traditional path of first performing spatiotemporal interpolation on the dynamic sea surface region to reconstruct a complete water surface elevation model, and then calculating the shoreline and climb rate accordingly. This application, by performing direct mapping statistics in the beam direction dimension and direct attribute discrimination in the point cloud dimension, fundamentally eliminates the dependence on spatiotemporal interpolation operations, achieving fast processing without interpolation. While ensuring processing accuracy, it significantly reduces computational overhead, meeting the needs of real-time detection.

[0095] In step S3, please refer to Figure 5In step S3, dynamic points are sliced ​​in time at preset time intervals, and shoreline candidate points are screened and spatial neighborhood constraints are applied to the dynamic points after time slicing to obtain the initial shoreline position. The initial shoreline position is then subjected to adaptive threshold correction to obtain time-series shoreline elevation data.

[0096] For example, please refer to Figure 9 Step S3 may include the following steps: Step S31 to Step S36.

[0097] Step S31: Perform time-series slicing on the dynamic points at preset time intervals to obtain multiple time-series slice dynamic point sets.

[0098] Step S32: Perform water depth threshold filtering on the time-series slice dynamic point set to obtain each initial shoreline candidate point set.

[0099] Step S33: Apply spatial neighborhood constraints to each initial shoreline candidate point set. From the candidate points that satisfy the spatial neighborhood constraints, select the point with the smallest coordinate value along the transverse shoreline direction and that satisfies the constraints on the number of dynamic points in the neighborhood and the consistency of elevation, to obtain the initial shoreline position and its three-dimensional coordinates for each time-series slice.

[0100] Step S34: Determine whether the distance between each initial shoreline position and the coastal protection structure is less than a preset distance threshold.

[0101] Step S35: If the distance is less than the preset distance threshold, the water depth threshold range is adaptively adjusted, and dynamic points are re-screened in the area in front of the coastal protection structure to obtain the corrected shoreline candidate point set and the corrected shoreline position.

[0102] Step S36: If the distance is greater than or equal to the preset distance threshold, the current shoreline position is taken as the final shoreline position. The final shoreline positions of all time-series slices are integrated in chronological order to obtain time-series shoreline elevation data covering the entire collection period.

[0103] As an example, in step S31, the preset time interval can be set to 0.6 seconds.

[0104] As an example, in step S32, dynamic points whose water depth values ​​are within the water depth threshold range are selected from the dynamic point set of each time slice to obtain the initial shoreline candidate point set.

[0105] As an example, the water depth threshold range can be set to a symmetrical or asymmetrical tolerance range that covers zero values.

[0106] In a preferred embodiment, when the z-axis in the local coastal coordinate system is vertically upward and positive, and the water depth value is defined as the dynamic point elevation minus the bed elevation, the water depth threshold range can be set to [-0.05m, 0.05m], meaning that dynamic points with water depth values ​​within 5 cm above and below zero are retained as candidate shoreline points. In shorelines with complex wave conditions or steep bed slopes, the water depth threshold range can be adjusted to [-0.1m, 0.1m]; under conditions of high point cloud signal-to-noise ratio and stable sea conditions, the water depth threshold range can also be reduced to [-0.03m, 0.03m] to improve shoreline spatial resolution. It is understood that specific values ​​can be calibrated based on on-site topographic survey data and wave observation statistics during the deployment of the detection system, and are not limited to the example values ​​mentioned above.

[0107] As an example, in step S33, a three-dimensional spatial neighborhood with a radius of R is defined in the local coastal coordinate system, centered on a candidate point in each initial shoreline candidate point set, as a spatial neighborhood constraint. The number of dynamic points within the spatial neighborhood constraint is counted. When the number is greater than or equal to the preset threshold N for the number of dynamic points in the neighborhood, the candidate point enters the next level of constraint judgment.

[0108] As an example, the preset threshold N for the number of dynamic neighborhood points can be set to 3.

[0109] Furthermore, the elevation standard deviation of all dynamic points within the spatial neighborhood constraint is calculated. When the elevation standard deviation is less than or equal to a preset elevation consistency threshold, the candidate point satisfies all the constraint conditions.

[0110] As an example, the preset elevation consistency threshold can be set to 0.2m.

[0111] Furthermore, from all candidate points that satisfy all constraints, the candidate point with the smallest coordinate value along the transverse shoreline, i.e. the candidate point closest to the landside, is selected as the initial shoreline position of the time slice, and its three-dimensional coordinates are obtained.

[0112] In one specific embodiment, a radius R = 2.0m can be set, a preset threshold for the number of dynamic neighborhood points N = 10, and a preset elevation consistency threshold of 0.10m can be set. For an initial shoreline candidate point, if there are 12 dynamic points within its 3D spatial neighborhood of R = 2.0m, and the elevation standard deviation of these dynamic points is 0.08m, then the candidate point passes all constraints; if there are only 3 dynamic points in the neighborhood, or the elevation standard deviation is 0.25m, then the candidate point is eliminated.

[0113] As an example, in step S34, the distance d between each initial shoreline position and the coastal protection structure is calculated, and it is determined whether the distance d is less than a preset distance threshold.

[0114] For example, the preset distance threshold is set to 2.0 meters. That is, when the initial shoreline position is less than 2.0 meters from the nearest coastal protection structure in the transverse direction, the initial shoreline position is considered to be close to the protection structure. In shorelines with large-scale protection structures and drastic topographic changes at the structure's leading edge, the preset distance threshold can be adjusted to 3.0 meters or 5.0 meters; in shorelines with compact structures and gentle slopes at the leading edge, the preset distance threshold can also be reduced to 1.0 meter. It is understood that the preset distance threshold can be calibrated during system deployment based on the measured geometric boundaries of the on-site coastal protection structure and the spatial resolution of the lidar point cloud, and is not limited to the above example value.

[0115] As an example, in step S35, if the distance d is less than the preset distance threshold, the lower limit of the water depth threshold range in step S32 is reduced, and the elevation constraint is relaxed. Dynamic points are re-screened in the area in front of the coastal protection structure to obtain a corrected shoreline candidate point set. Step S33 is repeated to obtain the corrected shoreline position, thus completing the shoreline position correction and avoiding interference from the coastal structure on the detection results.

[0116] Further, in step S36, if the distance d is greater than or equal to the preset distance threshold, then the current shoreline position is taken as the final shoreline position. The final shoreline positions of all time-series slices are integrated in chronological order to obtain continuous time-series shoreline elevation data covering the entire acquisition period.

[0117] In step S4, please refer to Figure 5 In step S4, wave run-up parameters are calculated based on the time-series shoreline elevation data. The wave run-up parameters include maximum run-up height, characteristic run-up height, and average shoreline elevation. When the maximum run-up height is greater than or equal to the top elevation of the coastal protection structure, it is determined to be a wave overrun, and the time and duration of the overrun are recorded.

[0118] Specifically, the local maximum value of shoreline elevation is extracted within a fixed time window as the single-wave climb height, and the maximum value among all single-wave climb heights is the maximum climb height within the fixed time window. All single-wave ascent heights are sorted from largest to smallest, and the values ​​with a cumulative probability exceeding 2% are taken as the feature ascent heights. Simultaneously, the arithmetic mean of all time-series shoreline elevation data within the fixed time window is calculated as the average shoreline elevation. .

[0119] As an example, the fixed time window is consistent with the duration of a single data collection.

[0120] Furthermore, the maximum climbing height Compare with the pre-obtained elevation of the top of the coastal protection structure. When the maximum climb height is... When the elevation is greater than or equal to the top elevation of the coastal protection structure, it is determined that a wave overtopping has occurred within the time window; the time of the overtopping is recorded, and monitoring continues until the maximum rise height is reached. The duration of the overtopping wave is obtained by measuring the drop below the top elevation of the coastal protection structure; and the maximum climb height is also recorded. The difference between the elevation of the seabed and the top elevation of the coastal protection structure is recorded as the overtopping height. When the maximum run-up height... If the elevation remains below the top elevation of the coastal protection structure, it is determined that no overtopping has occurred.

[0121] In step S5, please refer to Figure 5 In step S5, the wave climb parameters and overtopping detection results are output, and the corresponding synchronous video frames are automatically retrieved and output according to the maximum climb height or the key node where the overtopping occurs.

[0122] Specifically, the core quantitative data of the detection area are output in real time, including: beachbed topographic profile of the detection area, time-series shoreline location and elevation data, wave climb parameters, and the wave overrun detection results include wave overrun status, wave overrun time, wave overrun duration and wave overrun height.

[0123] As an example, when the maximum climbing height is detected When the elevation is greater than or equal to the top elevation of the coastal protection structure, it is determined that an overtopping has occurred, and this determination time is the critical node for the occurrence of an overtopping.

[0124] Furthermore, when it is determined that a wave overrun has occurred, the system automatically retrieves the corresponding synchronous video frame from the video data synchronously collected in step S1 using the time corresponding to the wave overrun event as an index, and outputs the synchronous video frame together with the corresponding quantitative detection data, thereby realizing the combined and linked output of quantitative detection data and qualitative visual information.

[0125] As an example, the processed detection data can be uploaded to the coastal disaster detection cloud platform in real time, supporting remote access, data query and visualization; at the same time, historical data backups are retained locally, and the backup cycle can be set according to storage capacity and traceability requirements to meet data traceability needs and realize local data storage and remote data upload functions.

[0126] As an example, the detection data can provide a quantitative basis for coastal disaster early warning. When the wave rise is detected to be close to the design threshold of the coastal protection structure or when overtopping occurs, the coastal management department can issue a disaster warning in a timely manner based on the detection data. At the same time, long-term wave rise and overtopping detection data can be used for performance evaluation, operation and maintenance optimization and coastal disaster risk assessment of coastal protection structures.

[0127] The wave detection method presented in this application exhibits strong environmental adaptability and high detection continuity. Through optimized point cloud classification and noise filtering algorithms, it effectively overcomes interference from rain, snow, steam, fog, and coastal protection structures on lidar data. It can operate stably at night, in adverse weather conditions, and under complex coastal terrain, achieving continuous, uninterrupted detection throughout the day, thus overcoming the environmental limitations of traditional optical detection. Furthermore, it boasts efficient data processing, meeting real-time detection requirements. It abandons the spatiotemporal interpolation operations of traditional lidar data processing, employing a direct point cloud data processing method, significantly simplifying the processing flow and reducing computational costs. The average processing time for a single data acquisition is far less than the acquisition duration, enabling real-time processing of detection data. The system provides timely output results to meet the requirements of coastal disaster early warning; the detection results are accurate and enable quantitative detection. In applications with different coastal types such as gravel beaches and mixed gravel beaches, the root mean square error of wave run-up measurement can be controlled within 0.35m, and the measurement error for mixed gravel beaches can be less than 0.2m. It can also accurately identify wave overtopping phenomena, making up for the shortcomings of traditional optical detection which can only perform qualitative observations, and providing quantitative data support for coastal disaster management. It can achieve continuous real-time quantitative detection of wave run-up height and wave overtopping phenomena in complex coastal topography and severe meteorological and hydrological conditions, providing data support for coastal disaster early warning, operation and maintenance of coastal protection projects, and coastal management.

[0128] It should be understood that although the steps in the flowcharts of the accompanying drawings are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some of the steps in the accompanying drawings may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.

[0129] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0130] Although this application has been disclosed above with reference to embodiments, it is not intended to limit this application. Anyone skilled in the art can make some modifications and refinements without departing from the spirit and scope of this application.

Claims

1. A method for detecting overtopping waves, characterized in that, Includes the following steps: Raw point cloud data is collected at preset time intervals. For each batch of raw point cloud data, the spatial angle of each beam emission direction is calculated. A beam angle grid is constructed at preset angle intervals, and the maximum detection distance of the scanning device within each grid is calculated to obtain the maximum distance mapping relationship of the beam direction. Based on the maximum distance mapping relationship of the beam direction, the original point cloud data is converted to the local coastal coordinate system and divided into grids. The beach bed elevation of each grid cell is determined. The point cloud data under the local coastal coordinate system is classified into static points and dynamic points. The water depth value of the dynamic points relative to the beach bed elevation is calculated. Dynamic points are sliced ​​in time at preset time intervals, and shoreline candidate points are screened and spatial neighborhood constraints are applied to the dynamic points after time-series slicing to obtain initial shoreline positions. Adaptive threshold correction is applied to the initial shoreline positions to obtain time-series shoreline elevation data. Wave run-up parameters are calculated based on the time-series shoreline elevation data, and the wave run-up parameters and wave overrun detection results are output.

2. The wave detection method according to claim 1, characterized in that, Determining the beachbed elevation of each grid cell includes: Based on the maximum distance mapping relationship of the beam direction, according to the beam emission direction corresponding to each grid cell, its maximum detection distance in the maximum distance mapping relationship of the beam direction is indexed; the elevation of the spatial point corresponding to the maximum detection distance is regarded as the candidate static target elevation of the grid cell; after traversing all grid cells, spatial filtering is performed on the candidate static target elevation to remove outliers and obtain the static elevation distribution of each grid cell; The elevation values ​​of all transformed point cloud data within each grid cell are extracted, and their statistical quantiles are calculated. The statistical quantiles are then compared and fused with the static elevation distribution to obtain the fused reference elevation. After neighborhood smoothing to eliminate local abnormal fluctuations, the beach bed elevation of each grid cell is obtained.

3. The wave detection method according to claim 2, characterized in that, Point cloud data in the local coastal coordinate system is classified into static points and dynamic points. The water depth values ​​of dynamic points relative to the beach bed elevation are calculated, including: Based on the static elevation distribution, the initial static elevation interval is formed by extending upward and downward by a first preset tolerance value. With the beach bed elevation as a reference, the lower boundary of the initial static elevation interval is constrained by a second preset tolerance value to obtain the reasonable range of the static target for each grid cell. Points whose elevation values ​​are within the reasonable range of the static target are determined as static points; the remaining points are determined as dynamic points.

4. The wave detection method according to claim 1, characterized in that, The process involves slicing dynamic points temporally at preset time intervals, filtering candidate shoreline points and constraining spatial neighborhoods on the sliced ​​dynamic points to obtain initial shoreline positions, and then applying adaptive threshold correction to these initial shoreline positions to obtain temporal shoreline elevation data, including: The dynamic points are processed by time-series slicing at preset time intervals to obtain multiple time-series slice dynamic point sets; The time-series slice dynamic point set is subjected to water depth threshold screening to obtain each initial shoreline candidate point set; Spatial neighborhood constraints are applied to each initial shoreline candidate point set. From the candidate points that satisfy the spatial neighborhood constraints, the point with the smallest coordinate value along the transverse shoreline direction and that satisfies the constraints on the number of dynamic points in the neighborhood and the consistency of elevation is selected to obtain the initial shoreline position and its three-dimensional coordinates for each time slice. Determine whether the distance between each initial shoreline location and the coastal protection structure is less than a preset distance threshold; If the distance is less than the preset distance threshold, the water depth threshold range is adaptively adjusted, and dynamic points are re-screened in the area in front of the coastal protection structure to obtain the corrected shoreline candidate point set and the corrected shoreline position. If the distance is greater than or equal to the preset distance threshold, the current shoreline position is taken as the final shoreline position. The final shoreline positions of all time-series slices are integrated in chronological order to obtain time-series shoreline elevation data covering the entire collection period.

5. The wave detection method according to claim 4, characterized in that, Spatial neighborhood constraints are applied to each initial shoreline candidate point set. From the candidate points that satisfy the spatial neighborhood constraints, the point with the smallest coordinate value along the transverse shoreline direction and that satisfies the constraints on the number of dynamic points in the neighborhood and the consistency of elevation is selected. This yields the initial shoreline position and its three-dimensional coordinates for each time-series slice, including: Taking any candidate point in each initial shoreline candidate point set as the center, a three-dimensional spatial neighborhood is defined in the local coastal coordinate system with a first radius as a spatial neighborhood constraint. The number of dynamic points within the spatial neighborhood constraint is counted. When the number is greater than or equal to a preset threshold for the number of dynamic points in the neighborhood, the candidate point enters the next level of constraint judgment. Calculate the elevation standard deviation of all dynamic points within the spatial neighborhood constraint. When the elevation standard deviation is less than or equal to a preset elevation consistency threshold, the candidate point satisfies all the constraint conditions. From all candidate points that satisfy all constraints, select the candidate point with the smallest coordinate value along the transverse bank direction as the initial shoreline position of the time slice, and obtain its three-dimensional coordinates.

6. The wave detection method according to claim 1, characterized in that, Also includes: Wave run-up parameters are calculated based on the time-series shoreline elevation data. The wave run-up parameters include maximum run-up height, characteristic run-up height, and average shoreline elevation. When the maximum run-up height is greater than or equal to the top elevation of the coastal protection structure, it is determined to be a wave overrun, and the time and duration of the overrun are recorded. Output wave climb parameters and wave overtopping detection results, and automatically retrieve and output corresponding synchronous video frames based on the maximum climb height or key nodes where wave overtopping occurs.

7. The wave detection method according to claim 6, characterized in that, The wave run-up parameters are calculated based on the time-series shoreline elevation data, including: The local maximum value of shoreline elevation is extracted using a fixed time window as the single-wave climb height, and the maximum value among all single-wave climb heights is the maximum climb height within the fixed time window. Sort all single-wave climbing heights from largest to smallest, and take the corresponding values ​​with a cumulative probability greater than 2% as the feature climbing heights; Calculate the arithmetic mean of all time-series shoreline elevation data within the fixed time window, and use it as the average shoreline elevation.

8. A wave-crossing detection system, characterized in that, include: The hardware deployment module is used to acquire raw point cloud data and video data of the detection area; The data synchronization acquisition module performs spatiotemporal synchronization acquisition of the original point cloud data and video data to obtain spatiotemporal synchronized point cloud data and spatiotemporal synchronized video data, and has an anomaly detection function. The lidar point cloud processing module adopts a direct point cloud processing method without spatiotemporal interpolation. It sequentially performs maximum distance detection, static and dynamic point cloud classification, and shoreline position detection on the spatiotemporal synchronized point cloud data to obtain the time-series shoreline elevation, and calculates wave run-up parameters and wave overtopping determination results accordingly. The detection result output module outputs the wave rise parameters and wave overshoot determination results, and uses the spatiotemporal synchronized video data linked to key nodes for visualization display, and stores the detection results locally and uploads them remotely.

9. The wave detection system according to claim 8, characterized in that, The lidar point cloud processing module adopts a direct point cloud data processing method that does not require spatiotemporal interpolation. It performs noise filtering, shoreline identification, and rapid calculation of wave rise and overtopping parameters through three core processes: maximum distance detection of beam direction, static and dynamic classification of point cloud data, and shoreline position detection and wave detection parameter calculation.

10. The wave detection system according to claim 8, characterized in that, The hardware deployment module includes: Coastal inspection pole, which serves as a supporting carrier; The scanning device is installed at a first preset height on the coastal detection pole. It is used to receive laser echo signals from the coastal detection area, perform active scanning and ranging processing, and output raw three-dimensional point cloud data. The monitoring equipment is installed at a second preset height on the coastal detection pole, and its field of view is completely overlapped with that of the scanning equipment after spatial calibration. It is used to receive visible light signals, perform lens distortion correction and video encoding processing, and output synchronous video stream data. The data processing terminal device is deployed at a third preset height on the coastal detection pole. It is used to receive the original three-dimensional point cloud data and synchronous video stream data, perform real-time data processing, parameter calculation and result storage, and remotely upload the final detection results to the coastal disaster detection cloud platform via wireless or wired means.

Images