Hydropower unmanned aerial vehicle river channel inspection ship identification method and device

Abstract

The present disclosure provides a kind of hydroelectric station unmanned aerial vehicle river channel inspection ship identification method and device, it is related to ship identification technical field, the method comprises: by unmanned aerial vehicle in specified area is flown, to collect multiple subzone laser point cloud data, wherein, specified area is divided into multiple subzone;Laser point cloud data is input to the three-dimensional point cloud detection network constructed in advance, to carry out feature extraction and detection segmentation, obtain the ship identification result in specified area;Based on ship identification result and the preset whitelist ship database, judge whether the foreign invading ship appears in specified area.It is thus, by detecting the laser point cloud data of unmanned aerial vehicle aerial photography, the ship in river channel is identified, by constructing the whitelist ship library of hydroelectric station, the features of the ship identified and whitelist ship library are compared, the allowed traffic ship of hydroelectric station can be investigated, and suspicious invading ship is accurately identified.

Description

Technical Field

[0001] This disclosure relates to the field of vessel identification technology, and in particular to a method and device for identifying vessels used by unmanned aerial vehicles (UAVs) for river patrol at hydropower stations. Background Technology

[0002] Safety is a critical aspect of hydropower station operation and maintenance, requiring frequent inspections of the upstream and downstream waterways, as well as the identification and location of suspicious intruding vessels, to ensure dam and vessel safety. However, given the several kilometers of waterway upstream and downstream, manual inspections are difficult to conduct daily and are time-consuming and labor-intensive.

[0003] Currently, some hydropower stations use drones for intelligent inspections, but these usually identify intruding vessels using visible light, which has a low accuracy rate. Summary of the Invention

[0004] This application proposes a method and device for identifying unmanned aerial vehicle (UAV) vessels used for river inspection at hydropower stations, aiming to at least partially solve one of the technical problems in related technologies.

[0005] The first aspect of this application proposes a method for identifying unmanned aerial vehicle (UAV) river patrol vessels used in hydropower stations, including:

[0006] The drone flies over a designated area to collect laser point cloud data from multiple partitions, wherein the designated area is divided into the multiple partitions;

[0007] The laser point cloud data is input into a pre-constructed 3D point cloud detection network for feature extraction and detection segmentation to obtain the ship identification results in the specified area.

[0008] Based on the vessel identification results and the preset whitelist vessel database, it is determined whether an intruding vessel has appeared in the designated area.

[0009] The second aspect of this application provides a device for identifying unmanned aerial vehicle (UAV) vessels used for river patrol at hydropower stations, comprising:

[0010] A collection module is used to collect laser point cloud data of multiple partitions by flying over a designated area using a drone; wherein the designated area is divided into the multiple partitions.

[0011] The identification module is used to input the laser point cloud data into a pre-constructed three-dimensional point cloud detection network for feature extraction and detection segmentation to obtain the ship identification result in the specified area;

[0012] The judgment module is used to determine whether an intruding foreign vessel has appeared in the specified area based on the vessel identification result and a preset whitelist vessel database.

[0013] A third aspect of this application provides an electronic device, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the hydropower station unmanned aerial vehicle (UAV) river inspection vessel identification method of this application.

[0014] The fourth aspect of this application provides a non-transitory computer-readable storage medium storing computer instructions for causing the computer to execute the method for identifying unmanned aerial vehicle (UAV) river patrol vessels for hydropower stations disclosed in the embodiments of this application.

[0015] In this embodiment, a drone first flies over a designated area to collect laser point cloud data from multiple partitions. The designated area is divided into multiple partitions. The laser point cloud data is then input into a pre-constructed 3D point cloud detection network for feature extraction and detection segmentation to obtain vessel identification results within the designated area. Based on the vessel identification results and a pre-defined whitelist of vessels, it is determined whether any intruding vessels are present in the designated area. Therefore, compared to visible light, this method offers higher identification accuracy, significantly improving the efficiency and reliability of intelligent inspection. By using a neural network to detect laser point cloud data captured by the drone, vessels in the river are identified. By constructing a whitelist of vessels for hydropower stations and comparing the identified vessels with the features in the whitelist, it is possible to screen vessels permitted to pass through the hydropower station and accurately identify suspicious intruding vessels.

[0016] Additional aspects and advantages of this disclosure will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this disclosure. Attached Figure Description

[0017] The above and / or additional aspects and advantages of this disclosure will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, in which:

[0018] Figure 1 This is a flowchart illustrating the method for identifying unmanned aerial vehicle (UAV) river patrol vessels for hydropower stations according to embodiments of this disclosure.

[0019] Figure 2 This is a schematic diagram of the structure of a three-dimensional point cloud detection network provided according to an embodiment of this disclosure;

[0020] Figure 3 This is a schematic diagram of a hydropower station unmanned aerial vehicle (UAV) river inspection vessel identification device according to an embodiment of this disclosure;

[0021] Figure 4This is a block diagram of an electronic device used to implement the method for identifying unmanned aerial vehicle (UAV) river patrol vessels in hydropower stations according to embodiments of this disclosure. Detailed Implementation

[0022] Embodiments of this disclosure are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are used only to explain this disclosure, and should not be construed as limiting this disclosure. Rather, embodiments of this disclosure include all variations, modifications, and equivalents falling within the spirit and scope of the appended claims.

[0023] It should be noted that the implementing entity of the hydropower station UAV river inspection vessel identification method in this embodiment can be a hydropower station UAV river inspection vessel identification device. This device can be implemented by software and / or hardware. This device can be configured in any electronic device with an image acquisition device, and there is no limitation here.

[0024] In this embodiment of the disclosure, the "hydropower station drone river inspection vessel identification device" will be used as the executing entity to perform the hydropower station drone river inspection vessel identification method, hereinafter referred to as "device" for description, and no limitation is made here.

[0025] Figure 1 This is a flowchart illustrating the method for identifying unmanned aerial vehicle (UAV) river patrol vessels at hydropower stations, provided in the first embodiment of this disclosure. Figure 1 As shown, the method includes:

[0026] S101: Use drones to fly over a designated area to collect laser point cloud data from multiple partitions, where the designated area is divided into multiple partitions.

[0027] Optionally, flight path planning parameters can be configured in the drone first, with each zone having corresponding flight path planning parameters. These parameters include yaw angle, shooting distance, flight path speed, and safe distance.

[0028] The drone is a multi-rotor aircraft equipped with visible light equipment and lidar.

[0029] Specifically, the terrain model of the upstream and downstream river channels of the hydropower station can be obtained first, and then drones and hangars can be installed near the dam. The drones are multi-rotor drones equipped with visible light and lidar, and the hangars support functions such as automatic battery swapping, charging, and vehicle changing for the drones.

[0030] This can be further divided into upstream, downstream, and midstream zones, etc., without further limitation. Alternatively, it can be divided into upstream and downstream zones, without further limitation.

[0031] The zones can be designated as cruise zones. It should be noted that drones can cruise based on different operating parameters when operating in different cruise zones.

[0032] Before the drone begins its patrol operation, operating parameters can be transmitted to the drone via a wired connection, or control information can be sent to the drone to configure its parameters.

[0033] Among them, the route planning parameters can be the parameters that the UAV needs to use when conducting patrol operations.

[0034] In the field of UAV aerial surveying or aerial photogrammetry, flight path planning parameters indeed involve several key factors to ensure that the UAV can complete its photography work accurately, efficiently, and safely during mission execution. Here are some important parameters that may be involved in flight path planning: 1. Yaw Angle: During aerial photography, the yaw angle refers to the angle between the UAV's longitudinal axis and its direction of travel. Adjusting the yaw angle helps control the camera's pointing, ensuring the camera lens is directly facing the target, which is especially important when performing photogrammetry or monitoring in a specific direction. 2. Capture Distance or Camera Range: This refers to the vertical or horizontal distance between the camera on the UAV and the ground target being photographed. This parameter determines the coverage area and image resolution, directly affecting the quality of the aerial survey results. 3. Flight Speed ​​Along the Route: Flight speed refers to the speed at which the UAV flies along a pre-set flight path. Appropriate speed settings not only affect the overall mission execution time but also relate to camera exposure time and image overlap, thus affecting the likelihood of generating high-quality 3D models or orthophotos in post-processing. 4. Safety Distance: Safety distance refers to the minimum distance that a UAV should maintain between itself and obstacles (such as buildings, power lines, trees, or other air traffic) during flight to ensure flight safety. During the flight path planning phase, factors such as terrain undulations and obstacle distribution must be considered to set a reasonable safety distance and avoid collision risks. In addition, flight path planning also needs to consider other parameters, such as altitude, ground sample distance (GSD), side-lap and forward-lap overlap, as well as flight path shape, start and end points, and turning radius, to ensure the successful completion of the flight mission and the acquisition of satisfactory aerial survey data.

[0035] S102: Input the laser point cloud data into a pre-built 3D point cloud detection network to perform feature extraction and detection segmentation, and obtain the ship identification results in the specified area.

[0036] Optionally, the initial 3D point cloud detection network can be trained first based on a pre-acquired ship point cloud model library to obtain a trained 3D point cloud detection network.

[0037] The 3D point cloud detection network includes a feature extraction module and a detection and segmentation module. The feature extraction module consists of multiple set abstraction layers, each of which contains a sampling layer, a grouping layer, and a point network layer.

[0038] It should be noted that the purpose of the feature extraction module is to take point cloud data as input, extract features from the point cloud data, and output the feature matrix of the point cloud data. This module consists of multiple ensemble abstraction layers, which are composed of three key layers: a sampling layer, a grouping layer, and a point network layer. The sampling layer aims to select a set of points from the input point cloud to represent the centroid of a local region. It uses an iterative farthest point sampling algorithm to select a subset from the given input points, and then finds the point farthest from other points in the subset. The grouping layer aims to construct a set of local regions by finding neighboring points around the centroid, and outputs the coordinates of each centroid point and its neighboring points. The point network layer aims to encode the local region pattern into a feature vector using a miniature point network. Each output local region consists of local features composed of its centroid and neighboring points.

[0039] Specifically, the detection and segmentation module aims to detect and segment the point clouds belonging to the ship portion of the point cloud data into ship point cloud models. This is achieved by merging and concatenating features from the back-interpolation upsampling and feature extraction modules to output point-wise category features, which are then used to determine whether each point belongs to the ship model.

[0040] Optionally, the drone flies along a predetermined route, its onboard lidar system emitting laser pulses and receiving the reflected signals. By calculating the time difference and angle information of the laser pulses' round trip, it can accurately acquire the three-dimensional coordinates of the water surface and the surface of the ships, forming dense point cloud data. Next, point cloud annotation can be performed. The collected raw point cloud data is manually refined using professional point cloud annotation tools to identify and mark the ship parts. For example, the ship's outline boundary is delineated, and key feature points or regions are marked for subsequent use in training machine learning or deep learning algorithms. Furthermore, a model library can be built. The annotated ship point clouds are organized and stored in a ship point cloud model library, with each ship corresponding to one or more manually annotated 3D point cloud models. These models can represent ship instances of various shapes, sizes, and attitudes, enriching the diversity of the training sample set. Finally, a 3D point cloud detection network can be trained using the constructed ship point cloud model library.

[0041] In this embodiment of the disclosure, a convolutional neural network (CNN) or a network structure specifically designed for point cloud data (such as PointNet, PointCNN, etc.) can be used to improve the automatic detection and recognition capability of ship targets in unknown point cloud data by learning from a large amount of labeled data.

[0042] It should be noted that when the drone is conducting inspections, the collected LiDAR data is input into the PointNet 3D point cloud detection network in a partitioned manner to detect and identify ship models in the river, output the point cloud model of the ship, and further determine whether it is an intruding foreign vessel.

[0043] Figure 2 This is a schematic diagram of the structure of a 3D point cloud detection network, such as... Figure 2 As shown, the system is divided into two parts: the upper part is the feature extraction module, and the lower part is the detection and segmentation module. In the feature extraction module, the point cloud data is processed through sampling layers, grouping layers, and point mesh layers to obtain the feature matrix. In the detection and segmentation module, the feature matrix undergoes upsampling, local feature processing, and merging / concatenation to finally obtain the detection result.

[0044] S103: Based on the vessel identification results and the preset whitelist vessel database, determine whether there are any foreign intrusion vessels in the specified area.

[0045] The target vessel is any vessel detected within a designated area. This designated area can be a river area, and is not limited thereto. The first feature matrix can be the feature matrix of the target vessel determined based on its laser point cloud data. The second feature matrix can be the feature matrix corresponding to any legitimate vessel in the whitelisted vessel database.

[0046] Among them, legitimate vessels can be those that are not foreign intruders.

[0047] The preset threshold can be a similarity threshold.

[0048] Optionally, if the vessel identification result indicates that a target vessel has been detected in a designated area, the vessel's laser point cloud data can be input into a 3D point cloud detection network to obtain the first feature matrix of the target vessel. Then, the similarity between the first feature matrix and the second feature matrices of each legitimate vessel in a preset whitelist vessel database can be determined. If the similarity between the first feature matrix of the target vessel and the second feature matrix of any legitimate vessel is greater than or equal to a preset threshold, the target vessel can be determined to be a legitimate vessel, and it can be determined that no intruding foreign vessels have appeared in the designated area.

[0049] It should be noted that a whitelist of vessels can be established to screen vessels permitted to pass through the hydroelectric power station. When a vessel model is detected, its similarity is compared with each feature matrix in the whitelist to determine whether it is an intruding vessel. Specifically, cosine similarity can be used to calculate the similarity.

[0050]

[0051] Optionally, if the similarity between the first feature matrix and each of the second feature matrices of the target vessel is less than a preset threshold, the target vessel information is sent to the regulatory user for identification.

[0052] If the target vessel is identified as an intruding foreign vessel, an early warning will be issued.

[0053] If the identification result indicates that the target vessel is a legitimate vessel, the first feature matrix is ​​stored in a pre-defined whitelist vessel database.

[0054] In this embodiment, a drone first flies over a designated area to collect laser point cloud data from multiple partitions. The designated area is divided into multiple partitions. The laser point cloud data is then input into a pre-constructed 3D point cloud detection network for feature extraction and detection segmentation to obtain vessel identification results within the designated area. Based on the vessel identification results and a pre-defined whitelist of vessels, it is determined whether any intruding vessels are present in the designated area. Therefore, compared to visible light, this method offers higher identification accuracy, significantly improving the efficiency and reliability of intelligent inspection. By using a neural network to detect laser point cloud data captured by the drone, vessels in the river are identified. By constructing a whitelist of vessels for hydropower stations and comparing the identified vessels with the features in the whitelist, it is possible to screen vessels permitted to pass through the hydropower station and accurately identify suspicious intruding vessels. In this embodiment of the disclosure, a PointNet-based 3D point cloud detection network is established and trained using a ship point cloud model library. When the UAV conducts inspections, the collected LiDAR data is input into the PointNet 3D point cloud detection network in a partitioned manner to detect and identify ship models in the river. If a ship model is detected, it is compared with the features of the whitelist ship library to determine whether it is an intruding ship. This can screen for ships allowed to pass through the hydropower station and accurately identify suspicious intruding ships.

[0055] Figure 3 This is a schematic diagram of a hydroelectric power station unmanned aerial vehicle (UAV) river inspection vessel identification device according to another embodiment of this disclosure. Figure 3 As shown, the UAV river patrol vessel identification device 300 for this hydropower station includes:

[0056] The collection module 310 is used to collect laser point cloud data of multiple partitions by flying over a designated area using a drone; wherein the designated area is divided into the multiple partitions.

[0057] The identification module 320 is used to input the laser point cloud data into a pre-constructed three-dimensional point cloud detection network for feature extraction and detection segmentation to obtain the ship identification result in the specified area;

[0058] The judgment module 330 is used to determine whether an intruding foreign vessel has appeared in the designated area based on the vessel identification result and a preset whitelist vessel database.

[0059] Optionally, the determination module is specifically used for:

[0060] If the vessel identification result indicates that a target vessel has been identified in the designated area, the vessel laser point cloud data of the target vessel is input into the three-dimensional point cloud detection network to obtain the first feature matrix of the target vessel.

[0061] Determine the similarity between the first feature matrix and the second feature matrix of each legitimate vessel in the preset whitelist vessel database;

[0062] If the similarity between the first feature matrix of the target vessel and the second feature matrix of any legitimate vessel is greater than or equal to a preset threshold, the target vessel is determined to be a legitimate vessel, and it is determined that no intruding vessel has appeared in the designated area.

[0063] Optionally, the determination module is further configured to:

[0064] If the similarity between the first feature matrix and each of the second feature matrices of the target vessel is less than the preset threshold, then the target vessel information is sent to the regulatory user for identification.

[0065] If the identification result indicates that the target vessel is an intruding foreign vessel, an early warning will be issued;

[0066] If the identification result indicates that the target vessel is a legitimate vessel, the first feature matrix is ​​stored in the preset whitelist vessel database.

[0067] Optionally, the collection module is further configured to:

[0068] The UAV is configured with flight path planning parameters, wherein each partition has corresponding flight path planning parameters;

[0069] The route planning parameters include yaw angle, shooting distance, route speed, and safety distance.

[0070] The drone is a multi-rotor type and is equipped with visible light equipment and lidar.

[0071] Optionally, the identification module is further configured to:

[0072] The initial 3D point cloud detection network is trained based on a pre-acquired ship point cloud model library to obtain a trained 3D point cloud detection network.

[0073] The 3D point cloud detection network includes a feature extraction module and a detection and segmentation module. The feature extraction module consists of multiple set abstraction layers, each of which contains a sampling layer, a grouping layer, and a point network layer.

[0074] In this embodiment, a drone first flies over a designated area to collect laser point cloud data from multiple partitions. The designated area is divided into multiple partitions. The laser point cloud data is then input into a pre-constructed 3D point cloud detection network for feature extraction and detection segmentation to obtain vessel identification results within the designated area. Based on the vessel identification results and a pre-defined whitelist of vessels, it is determined whether any intruding vessels are present in the designated area. Therefore, compared to visible light, this method offers higher identification accuracy, significantly improving the efficiency and reliability of intelligent inspection. By using a neural network to detect laser point cloud data captured by the drone, vessels in the river are identified. By constructing a whitelist of vessels for hydropower stations and comparing the identified vessels with the features in the whitelist, it is possible to screen vessels permitted to pass through the hydropower station and accurately identify suspicious intruding vessels.

[0075] According to embodiments of this disclosure, this disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0076] Figure 4 A block diagram of an exemplary computer device suitable for implementing embodiments of the present application is shown. Figure 4 The computer device 12 shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments of this application.

[0077] like Figure 4 As shown, the computer device 12 is represented in the form of a general-purpose computing device. The components of the computer device 12 may include, but are not limited to: one or more processors or processing units 16, system memory 28, and a bus 18 connecting different system components (including system memory 28 and processing unit 16).

[0078] Bus 18 represents one or more of several bus architectures, including a memory bus or memory controller, a peripheral bus, a graphics acceleration port, a processor, or a local bus using any of the various bus architectures. Examples of these architectures include, but are not limited to, the Industry Standard Architecture (ISA) bus, the Micro Channel Architecture (MAC) bus, the Enhanced ISA bus, the Video Electronics Standards Association (VESA) local bus, and the Peripheral Component Interconnect (PCI) bus.

[0079] Computer device 12 typically includes a variety of computer system readable media. These media can be any available media that can be accessed by computer device 12, including volatile and non-volatile media, removable and non-removable media.

[0080] Memory 28 may include computer system readable media in the form of volatile memory, such as Random Access Memory (RAM) 30 and / or cache memory 32. Computer device 12 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, storage system 34 may be used to read and write non-removable, non-volatile magnetic media (…). Figure 4 Not shown; usually referred to as a "hard drive".

[0081] although Figure 4 Not shown, a disk drive for reading and writing to a removable non-volatile disk (e.g., a "floppy disk") and an optical disc drive for reading and writing to a removable non-volatile optical disc (e.g., a compact disc read-only memory (CD-ROM), a digital video disc read-only memory (DVD-ROM), or other optical media) may be provided. In these cases, each drive may be connected to bus 18 via one or more data media interfaces. Memory 28 may include at least one program product having a set (e.g., at least one) of program modules configured to perform the functions of the embodiments of this application.

[0082] A program / utility 40 having a set (at least one) of program modules 42 may be stored, for example, in memory 28. Such program modules 42 include, but are not limited to, an operating system, one or more application programs, other program modules, and program data. Each or some combination of these examples may include an implementation of a network environment. Program modules 42 typically perform the functions and / or methods described in the embodiments of this application.

[0083] Computer device 12 can also communicate with one or more external devices 14 (e.g., keyboard, pointing device, display 24, etc.), and with one or more devices that enable a user to interact with computer device 12, and / or with any device that enables computer device 12 to communicate with one or more other computing devices (e.g., network card, modem, etc.). This communication can be performed via input / output (I / O) interface 22. Furthermore, computer device 12 can also communicate with one or more networks (e.g., local area network (LAN), wide area network (WAN), and / or public networks, such as the Internet) via network adapter 20. As shown, network adapter 20 communicates with other modules of computer device 12 via bus 18. It should be understood that, although not shown in the figures, other hardware and / or software modules can be used in conjunction with computer device 12, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0084] The processing unit 16 executes various functional applications and data processing by running programs stored in the system memory 28, such as implementing the hydropower station drone river inspection vessel identification method mentioned in the foregoing embodiments.

[0085] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the application disclosed herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this application are indicated by the following claims.

[0086] It should be understood that this application is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this application is limited only by the appended claims.

[0087] It should be noted that in the description of this application, the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance. Furthermore, in the description of this application, unless otherwise stated, "a plurality of" means two or more.

[0088] Any process or method described in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing a particular logical function or process, and the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the function involved, as will be understood by those skilled in the art to which embodiments of this application pertain.

[0089] It should be understood that various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0090] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0091] Furthermore, the functional units in the various embodiments of this application can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. The storage medium mentioned above can be a read-only memory, a disk, or an optical disk, etc.

[0092] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0093] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. A method for identifying unmanned aerial vehicle (UAV) patrol vessels used in hydropower stations, characterized in that, include: The drone flies over a designated area to collect laser point cloud data from multiple partitions, wherein the designated area is divided into the multiple partitions; The initial 3D point cloud detection network is trained based on a pre-acquired ship point cloud model library to obtain a trained 3D point cloud detection network. The 3D point cloud detection network includes a feature extraction module and a detection and segmentation module. The feature extraction module consists of multiple set abstraction layers, each of which includes a sampling layer, a grouping layer, and a dot mesh layer. The laser point cloud data is input into the three-dimensional point cloud detection network for feature extraction and detection segmentation to obtain the ship identification result in the specified area; If the vessel identification result indicates that a target vessel has been identified in the designated area, the vessel laser point cloud data of the target vessel is input into the three-dimensional point cloud detection network to obtain the first feature matrix of the target vessel. Determine the similarity between the first feature matrix and the second feature matrix of each legitimate vessel in the preset whitelist vessel database; If the similarity between the first feature matrix of the target vessel and the second feature matrix of any legitimate vessel is greater than or equal to a preset threshold, the target vessel is determined to be a legitimate vessel, and it is determined that no intruding vessel has appeared in the designated area. After determining the similarity between the first feature matrix and the second feature matrix of each legitimate vessel in the preset whitelist vessel database, the method further includes: If the similarity between the first feature matrix and each of the second feature matrices of the target vessel is less than the preset threshold, then the target vessel information is sent to the regulatory user for identification. If the identification result indicates that the target vessel is an intruding foreign vessel, an early warning will be issued; If the identification result indicates that the target vessel is a legitimate vessel, the first feature matrix is ​​stored in the preset whitelist vessel database.

2. The method according to claim 1, characterized in that, Prior to the aforementioned drone flight over a designated area, the following is also included: The UAV is configured with flight path planning parameters, wherein each partition has corresponding flight path planning parameters; The route planning parameters include yaw angle, shooting distance, route speed, and safety distance. The drone is a multi-rotor type and is equipped with visible light equipment and lidar.

3. A device for identifying unmanned aerial vehicle (UAV) vessels used for river patrol at hydropower stations, characterized in that, include: A collection module is used to collect laser point cloud data of multiple partitions by flying over a designated area using a drone; wherein the designated area is divided into the multiple partitions. The identification module is used to train an initial 3D point cloud detection network based on a pre-acquired ship point cloud model library to obtain a trained 3D point cloud detection network. The 3D point cloud detection network includes a feature extraction module and a detection and segmentation module. The feature extraction module consists of multiple ensemble abstraction layers, each containing a sampling layer, a grouping layer, and a dot matrix layer. The laser point cloud data is input into the 3D point cloud detection network for feature extraction and detection and segmentation to obtain the ship identification result in the specified area. The judgment module is used to, when the vessel identification result indicates that a target vessel has been identified in the designated area, input the vessel laser point cloud data of the target vessel into the three-dimensional point cloud detection network to obtain a first feature matrix of the target vessel; determine the similarity between the first feature matrix and the second feature matrices of each legitimate vessel in a preset whitelist vessel database; and, if the similarity between the first feature matrix of the target vessel and the second feature matrix of any legitimate vessel is greater than or equal to a preset threshold, determine that the target vessel is a legitimate vessel and determine that no intruding vessel has appeared in the designated area. The judgment module is also used for: If the similarity between the first feature matrix and each of the second feature matrices of the target vessel is less than the preset threshold, then the target vessel information is sent to the regulatory user for identification. If the identification result indicates that the target vessel is an intruding foreign vessel, an early warning will be issued; If the identification result indicates that the target vessel is a legitimate vessel, the first feature matrix is ​​stored in the preset whitelist vessel database.

4. The apparatus according to claim 3, characterized in that, The collection module is also used for: The UAV is configured with flight path planning parameters, wherein each partition has corresponding flight path planning parameters; The route planning parameters include yaw angle, shooting distance, route speed, and safety distance. The drone is a multi-rotor type and is equipped with visible light equipment and lidar.

Images