Unmanned aerial vehicle laser obstacle removing system and method based on visual method

Abstract

The invention relates to the technical field of unmanned aerial vehicles, and discloses an unmanned aerial vehicle laser obstacle removing system and method based on a visual method, and the system is integrated with a high-definition camera and an advanced image processing unit, and carries out the real-time capturing and analysis of images of a power transmission line and the surrounding environment of the power transmission line through the deep learning technology. According to the method, the type, the position, the size and the shape of the obstacle can be accurately recognized and converted into accurate target positioning information, reliable guidance is provided for laser obstacle clearing operation, the obstacle recognition and positioning precision is remarkably improved through the improvement, and the accuracy and the efficiency of laser obstacle clearing are ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicles (UAVs), and in particular to a UAV laser obstacle removal system and method based on a vision method. Background Art

[0002] Drone laser obstacle removal technology is a significant innovation in modern power line inspection and maintenance. This technology utilizes high-precision laser equipment aboard drones to remotely and contactlessly remove obstacles from power transmission lines. Compared to traditional manual obstacle removal methods, drone laser obstacle removal offers advantages such as high efficiency, low safety risks, and wide applicability. It can quickly and accurately locate and remove obstacles such as tree branches, plastic bags, and bird nests in complex and changing environments, effectively ensuring the safe and stable operation of power transmission lines.

[0003] Although UAV laser obstacle removal technology has made significant progress, there are still some problems and challenges in practical application. For example:

[0004] 1. Existing drone laser obstacle removal systems have deficiencies in obstacle recognition and positioning. Due to the complex and ever-changing environment surrounding power transmission lines and the wide variety of obstacles with varying shapes, traditional image processing methods often have difficulty accurately identifying and locating these obstacles, limiting the accuracy and efficiency of laser obstacle removal.

[0005] 2. The stability of drones in complex flight environments is also an important factor restricting the effectiveness of laser obstacle removal. External factors such as wind and airflow may cause the drone's posture to change, thereby affecting the camera's locking stability and the laser beam's emission accuracy, increasing the risk and uncertainty of obstacle removal operations.

[0006] In order to solve the above problems, this application proposes a UAV laser obstacle removal system and method based on visual methods. Summary of the Invention

[0007] The present invention provides a UAV laser obstacle removal system and method based on a vision method to solve the problems existing in the prior art.

[0008] In order to achieve the above object, the present invention is implemented through the following technical solutions:

[0009] In a first aspect, the present invention provides a UAV laser obstacle removal system based on a visual method, comprising:

[0010] The visual recognition module is used to identify obstacle information and generate target positioning information based on the obstacle information;

[0011] The UAV laser obstacle removal module is used to emit a laser beam to cut and remove obstacles on the power transmission line based on the target positioning information provided by the visual recognition module;

[0012] An anti-shake locking module for controlling the stable locking of obstacles by the visual recognition module in a complex flight environment;

[0013] A ground control terminal for sending control instructions to the visual recognition module, and also for real-time monitoring of the UAV's status, remotely controlling the UAV to perform obstacle removal tasks, and collecting and analyzing task data.

[0014] Optionally, the visual recognition module includes a visual camera, a lidar, a spectrometer, and a thermal imaging sensor;

[0015] The visual camera is used for capturing, processing, and analyzing obstacle image information in real time after receiving control instructions from the ground control terminal, and is also used for calculating the three-dimensional coordinates of obstacles based on the obstacle image information and the lidar, and identifying the material type through the spectrometer and the laser-induced breakdown spectroscopy probe, and by laser reflection spectroscopy;

[0016] The thermal imaging sensor is used for identifying live wires and humans by fusing RGB, LiDAR, and thermal imaging data through a deep learning model, and optimizing the obstacle information and target positioning information based on the identification results.

[0017] Optionally, the visual camera includes a high-definition camera and an image processing unit. The high-definition camera is used for capturing real-time images of the transmission line and the surrounding environment; the image processing unit is used for preprocessing, feature extraction, obstacle recognition, and target positioning of the real-time images to obtain target positioning information.

[0018] Optionally, the obstacle information includes the size and material of the obstacle, and the UAV laser obstacle removal module includes:

[0019] A multi-laser emission unit for configuring laser emitters with different powers and wavelengths according to the size and material of the obstacle;

[0020] A laser automatic aiming module for aiming the multi-laser emission unit at the target according to the target information to guide the emission of the laser beam;

[0021] A stable flight platform equipped with GPS navigation, an inertial navigation system, and obstacle avoidance sensors to ensure the stable flight of the UAV in a complex environment.

[0022] Optionally, the anti-shake locking module includes:

[0023] A center of gravity adjustment and stabilization unit provided with a movable counterweight for balancing the fuselage in real time according to the movable counterweight;

[0024] An automatic anti-shake unit, built-in with a gyroscope and an accelerometer, is used to monitor the attitude changes of the drone in real time and automatically adjust the image capture attitude to maintain a stable lock on obstacles.

[0025] Optionally, the anti-shake locking module is specifically a multi-degree-of-freedom flexible gimbal. The structural type of the multi-degree-of-freedom flexible gimbal is a flexible damping material or a non-linear spring structure, including an S-shaped multi-segment structure. Each adjacent two segments are connected by a joint. Each joint is built-in with a micro hydraulic damping cavity. Each joint is equipped with a piezoelectric fiber sensor for monitoring deformation in real time and feeding back to the control unit. A micro pneumatic artificial muscle is used, and rapid contraction or extension is achieved through air pressure changes. The driving layout is specifically arranged with 4 groups of artificial muscles in a cross arrangement to form an antagonistic structure similar to biological muscles; the robotic arm of the multi-degree-of-freedom flexible gimbal uses a foldable pneumatic wind-resistant robotic arm. A vortex generator is set at the folding joint to disrupt the separated air flow to reduce turbulence. The active wind-resistant structure is set that a micro jet nozzle is integrated at the end of the robotic arm, and high-speed air flow is ejected according to the data of the air pressure sensor to actively offset the side wind moment; the multi-degree-of-freedom flexible gimbal specifically adopts a piezoelectric-electromagnetic hybrid energy recovery gimbal. The structural design adopts a vibration energy recovery layer. A piezoelectric ceramic sheet array is embedded in the gimbal bracket to convert mechanical vibration into electrical energy. A Hall electromagnetic coil group is superimposed, and electricity is generated by the gimbal swinging to cut the magnetic induction line, forming a complement with the piezoelectric effect.

[0026] Optionally, the ground control terminal includes:

[0027] A real-time monitoring interface unit, used to display the real-time video stream, flight parameters, obstacle recognition results and laser obstacle clearing progress of the drone;

[0028] A remote control unit, used to send control commands to the vision recognition module to remotely command the takeoff, landing and route planning of the drone, and also used to send obstacle clearing commands to the drone laser obstacle clearing module to perform laser obstacle clearing operations.

[0029] Optionally, the ground control terminal further includes:

[0030] A data analysis and reporting unit, responsible for automatically recording the historical data of the obstacle clearing task and generating reports for subsequent analysis and optimization.

[0031] This application also provides a method for a drone to clear obstacles with a laser based on a vision method, including:

[0032] S1: When the system starts, each module is initialized to ensure that they are in a normal working state;

[0033] S2: The ground control terminal displays the system status. After the operator confirms that it is correct, the task is started;

[0034] S3: The visual recognition module starts to capture and process images, identify obstacles, and transmit target location information.

[0035] S4: According to the target location information, the UAV laser obstacle clearing module selects a laser emitter to work and accurately emits a laser beam to cut the obstacle.

[0036] S5: The anti-shake locking module ensures the stable locking of the camera on the obstacle in a complex flight environment, providing accurate target information for laser obstacle clearing.

[0037] S6: The ground control terminal monitors the status of the UAV, the progress of obstacle clearing, and other information in real time, and performs remote control as needed.

[0038] S7: After the obstacle clearing task is completed, the UAV returns to the base or a designated location.

[0039] S8: The ground control terminal collects and analyzes the task data, generating a detailed report for subsequent analysis and optimization.

[0040] Beneficial effects:

[0041] The UAV laser obstacle clearing system based on the vision method provided by the present invention integrates a high-definition camera and an advanced image processing unit, and uses deep learning technology to capture and analyze images of the transmission line and its surrounding environment in real time. It can accurately identify the type, location, size, and shape of obstacles, and convert them into accurate target location information, providing reliable guidance for laser obstacle clearing operations. This improvement significantly improves the accuracy of obstacle recognition and location, ensuring the accuracy and efficiency of laser obstacle clearing.

[0042] In a further technical solution, through the design of the anti-shake locking module, the stable locking of the camera on the obstacle in a complex flight environment is achieved. The coordinated work of the automatic anti-shake unit and the intelligent algorithm unit can monitor the attitude change of the UAV in real time and automatically adjust the locking position of the camera, maintaining the accurate alignment between the laser beam and the obstacle. At the same time, the installation of a stable flight platform also ensures the stable flight of the UAV in a complex environment, providing a stable working platform for laser obstacle clearing.

[0043] In a further technical solution, through the data analysis and reporting unit of the ground control terminal, the automatic recording and analysis of the historical data of the UAV obstacle clearing task are realized. These data include key information such as obstacle type, location, and obstacle clearing efficiency, which can provide valuable references for subsequent task planning and execution strategies. By continuously optimizing the task data and algorithm model, the overall performance and reliability of the UAV laser obstacle clearing system can be further improved, providing a more powerful guarantee for the stable operation of the power system. Description of the drawings

[0044] Figure 1 Block diagram of a UAV laser obstacle clearing system based on a vision method according to a preferred embodiment of the present invention;

[0045] Figure 2 Flowchart of a UAV laser obstacle clearing method based on a vision method according to a preferred embodiment of the present invention. Detailed implementation manners

[0046] The technical solutions of the present invention will be described clearly and completely below. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0047] Unless otherwise defined, the technical terms or scientific terms used in the present invention shall have the ordinary meanings understood by those of ordinary skill in the art to which the present invention belongs. The "first", "second" and similar terms used in the present invention do not denote any order, quantity or importance, but are only used to distinguish different components. Similarly, terms such as "one" or "a" do not denote a quantity limitation, but mean that there is at least one. The term "connected" or "coupled" and the like are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The terms "upper", "lower", "left", "right" and the like are only used to represent relative positional relationships, and when the absolute position of the object being described changes, the relative positional relationship also changes accordingly.

[0048] Please refer to Figure 1 , a UAV laser obstacle clearing system based on a vision method provided by the present application, includes:

[0049] A vision recognition module, configured to recognize obstacle information and generate target positioning information according to the obstacle information;

[0050] A UAV laser obstacle clearing module, configured to emit a laser beam to cut and clear obstacles on a power transmission line according to the target positioning information provided by the vision recognition module;

[0051] An anti-shake locking module, configured to control the vision recognition module to stably lock on an obstacle in a complex flight environment;

[0052] A ground control terminal, configured to send control instructions to the vision recognition module, and also configured to monitor the state of the UAV in real time, remotely control the UAV to perform an obstacle clearing task, and collect and analyze task data.

[0053] [[ID=O]]In this embodiment, the visual recognition module uses multi-sensor feature recognition and positioning, including a visual camera, a lidar, a spectrometer, and a thermal imaging sensor. The visual camera is responsible for capturing, processing, and analyzing obstacle image information in real time, and combines with the lidar to accurately calculate the three-dimensional coordinates of the obstacle. The spectrometer (wavelength range 400-1100nm, resolution 1nm) and the laser-induced breakdown spectroscopy (LIBS) probe identify the material type through laser reflection spectroscopy to determine the obstacle type. The thermal imaging sensor is used to identify live wires and humans. By fusing RGB, LiDAR, and thermal imaging data through a deep learning model, the recognition accuracy in complex environments (such as backlighting and foggy days) is improved.

[0054] The above-mentioned UAV laser obstacle clearing system based on the visual method integrates a high-definition camera and an advanced image processing unit, and uses deep learning technology to capture and analyze images of the transmission line and its surrounding environment in real time. It can accurately identify the type, position, size, and shape of the obstacle, and convert it into accurate target positioning information, providing reliable guidance for the laser obstacle clearing operation. This improvement significantly improves the accuracy of obstacle recognition and positioning, ensuring the accuracy and efficiency of laser obstacle clearing.

[0055] Optionally, the visual camera includes a high-definition camera and an image processing unit. The high-definition camera is used to capture real-time images of the transmission line and its surrounding environment; the image processing unit is used to preprocess the real-time images, extract features, identify obstacles, and perform target positioning to obtain target positioning information.

[0056] In this optional embodiment, the high-definition camera and the image processing unit work together to achieve efficient recognition and positioning of transmission line obstacles. By using advanced image processing and deep learning technologies, it can accurately identify the type, position, size, and shape of obstacles in complex and changing environments, providing precise guidance for the laser obstacle clearing operation of the UAV. The specific process is as follows:

[0057] High-definition camera startup and initialization: After the device is powered on or receives a startup instruction, the high-definition camera performs self-checks, including checking the integrity of the lens, the normal connection of the sensor, etc. During the initialization process, appropriate parameters such as resolution and frame rate are set to ensure that image data that meets the requirements can be obtained. Image capture: According to the set frame rate of 30 frames per second, the high-definition camera continuously captures images of the surrounding environment. Light enters the camera through the lens and is projected onto the image sensor CMOS. The sensor converts the optical signal into an analog electrical signal, which is then converted into a digital image signal by an analog-to-digital converter (ADC). The high-definition camera adopts a high-resolution, wide-angle lens design. Based on advanced image sensor and image processing technologies, it has excellent image capture capabilities. It can automatically adjust parameters such as exposure and white balance to meet the shooting requirements under different light conditions, ensuring clear and delicate images of the transmission line and its surrounding environment are captured;

[0058] The image processing unit incorporates deep learning algorithms and can analyze the image data captured by the high-definition camera in real time. The specific workflow is as follows:

[0059] 1. Image preprocessing: Preprocess the original images captured by the high-definition camera, including operations such as denoising and enhancing contrast, to improve the image quality and provide reliable input for subsequent analysis;

[0060] 2. Feature extraction: Use deep learning algorithms to extract features from the preprocessed images, including information such as the edges, textures, and colors of obstacles, which helps to identify the types and shapes of obstacles;

[0061] 3. Obstacle recognition: Based on the extracted features, the deep learning algorithm further recognizes the obstacles in the images. By training a large number of sample data, the algorithm can accurately identify obstacles made of different materials such as branches, plastics, and metals, and determine their positions, sizes, and shapes;

[0062] 4. Target localization: Convert the recognized obstacle information into precise target localization information, including the center coordinates and bounding boxes of the obstacles. This information will be transmitted to the UAV laser obstacle removal module to guide the precise emission of the laser beam;

[0063] In summary, the visual recognition module constitutes the core part of the UAV laser obstacle removal system. By capturing and analyzing images of the transmission line and its surrounding environment in real time, it provides precise target localization information for subsequent laser obstacle removal operations. During this process, the high-definition camera ensures clear image capture, while the image processing unit uses deep learning algorithms to achieve accurate identification and localization of obstacles. The two complement each other and jointly promote the efficient operation of the UAV laser obstacle removal system.

[0064] Optionally, the UAV laser obstacle removal module includes:

[0065] A multi - laser emission unit, which is used to configure laser emitters with different powers and wavelengths according to the size and material of obstacles;

[0066] A laser automatic aiming module, which is used to aim the multi - laser emission unit at the target according to the target information to guide the emission of laser beams;

[0067] A stable flight platform, which is equipped with GPS navigation, an inertial navigation system and obstacle avoidance sensors to ensure the stable flight of the UAV in complex environments.

[0068] In this alternative embodiment, the UAV laser obstacle clearing module integrates a multi - laser emission unit and a stable flight platform. Through the joint cooperation of the multi - laser emission unit and the stable flight platform, and according to the target positioning information provided by the visual recognition module, laser beams are accurately emitted to cut and clear obstacles on the transmission line. The specific working content is as follows:

[0069] The multi - laser emission unit configures a series of laser emitters with different powers and wavelengths according to the different sizes and materials of obstacles. The working principles of these laser emitters are as follows:

[0070] 1. Laser beam emission: When receiving the target positioning information from the visual recognition module, the multi - laser emission unit will select the most suitable laser emitter to work according to the type and size of the obstacle, and the laser beam is accurately aimed at the obstacle to ensure the cutting effect;

[0071] 2. Power and wavelength adjustment: Different obstacles respond differently to lasers. For example, branches may require lasers with higher power to cut, while light materials such as plastic bags may require lasers with lower power to avoid over - burning. The multi - laser emission unit adjusts the power and wavelength of the laser to adapt to the cutting requirements of different obstacles;

[0072] 3. Safety protection mechanism: To ensure the safety of operations, the multi - laser emission unit has built - in multiple safety protection mechanisms. For example, when it is detected that the laser beam may be misdirected to the transmission line or other important facilities, the system will automatically turn off the laser emitter to avoid potential risks;

[0073] The stable flight platform can ensure the stable flight of the UAV in complex environments and provide a stable working platform for laser obstacle clearing. The specific working principle is as follows:

[0074] 1. UAV control technology: The stable flight platform adopts advanced UAV control technologies such as GPS navigation, inertial navigation system (INS) and obstacle avoidance sensors. Through the joint cooperation of these technologies, the stability and accuracy of the UAV during flight are ensured;

[0075] 2. Flight attitude adjustment: During the flight of the drone, it may be affected by external factors such as wind and air currents, resulting in changes in the flight attitude. The stable flight platform monitors the attitude changes of the drone in real time and automatically adjusts the flight attitude to ensure that the drone can maintain stable flight;

[0076] 3. Obstacle avoidance function: To prevent the drone from colliding with obstacles during flight, the stable flight platform is equipped with advanced obstacle avoidance sensors that can monitor the environment around the drone in real time and automatically adjust the flight route when potential obstacles are detected to ensure the safety of the drone;

[0077] 4. Laser obstacle clearing work platform: While the drone is flying stably, the stable flight platform also provides a stable work platform for laser obstacle clearing. The laser automatic aiming module aims the multi-laser emission unit at the target location processed by the visual recognition module to guide the precise emission of the laser beam, which ensures that the laser beam can accurately aim at the obstacle and achieve an effective cutting effect;

[0078] In summary, the drone laser obstacle clearing module and its multiple units cooperate together to efficiently and safely clear the obstacles on the transmission line. During this process, the multi-laser emission unit precisely emits laser beams for cutting according to the different sizes and materials of the obstacles, while the stable flight platform ensures the stable flight of the drone in a complex environment and provides a stable work platform for laser obstacle clearing. Through the close cooperation of these two components, the drone laser obstacle clearing system can efficiently and accurately complete the obstacle clearing task in various complex environments.

[0079] Optionally, the anti-shake locking module includes:

[0080] The center of gravity adjustment and stabilization unit is provided with a movable counterweight block for balancing the fuselage in real time according to the movable counterweight block;

[0081] The automatic anti-shake unit is built-in with a gyroscope and an accelerometer for monitoring the attitude changes of the drone in real time and automatically adjusting the image capture attitude to maintain a stable lock on the obstacle.

[0082] In this optional embodiment, the anti-shake locking module integrates a multi-degree-of-freedom flexible gimbal, a center of gravity adjustment and stabilization unit, and an automatic anti-shake unit. Through the cooperation of multiple units, it ensures the stable lock of the camera on the obstacle in a complex flight environment and provides accurate target information for subsequent laser obstacle clearing operations. The specific working content is as follows:

[0083] The multi-degree-of-freedom flexible gimbal incorporates flexible damping materials (such as silicone-carbon fiber composite layers, titanium alloy-silicone composite hinges) or non-linear spring structures, mimicking the S-shaped multi-segment structure of bird cervical vertebrae. Each joint is equipped with a micro hydraulic damping chamber (filled with magnetorheological fluid), and each joint is equipped with a piezoelectric fiber sensor (embedded in the silicone layer) to monitor deformation in real time and feedback it to the control unit.

[0084] Use a micro pneumatic artificial muscle (diameter 3mm, nylon braided tube + pneumatic drive) to replace the traditional motor, achieving rapid contraction or extension through air pressure changes, with a response time less than 10ms.

[0085] The drive layout specifically arranges 4 groups of artificial muscles in a cross pattern to form an antagonistic structure similar to biological muscles.

[0086] It should be noted that when the drone is impacted by a crosswind, the pneumatic muscle quickly contracts to adjust the gimbal attitude, and at the same time, the magnetorheological fluid instantaneously thickens under the action of the electromagnetic field to absorb high-frequency vibrations, such as the vibration of the propeller.

[0087] It should be noted that the gimbal specifically adopted is a piezoelectric-electromagnetic hybrid energy recovery gimbal. The structural design uses a vibration energy recovery layer. Piezoelectric ceramic sheet arrays (PZT-5H material) are embedded in the gimbal bracket to convert mechanical vibration into electrical energy (a single vibration pulse can generate 5V / 10mA), and Hall electromagnetic coil groups are superimposed to generate electricity by using the gimbal swing to cut the magnetic induction line, forming a complement to the piezoelectric effect.

[0088] The self-power supply circuit stores the recovered electrical energy in a super capacitor and directly powers the gimbal sensors (such as gyroscopes, encoders), reducing the load on the main battery.

[0089] Center of gravity adjustment and stabilization unit: Balance the fuselage in real time through a movable counterweight to reduce the attitude deviation caused by rapid acceleration.

[0090] The automatic anti-shake unit is built-in with high-precision gyroscopes and accelerometers, which can monitor the attitude changes of the drone in real time. When the drone is affected by external factors such as wind and air currents during flight, its attitude may change, resulting in unstable locking of the camera on the obstacle. At this time, the automatic anti-shake unit will respond quickly and automatically adjust the attitude of the camera according to the data provided by the gyroscopes and accelerometers to maintain stable locking on the obstacle. This real-time adjustment ability ensures that even in a complex and changeable flight environment, the camera can always accurately lock on the target obstacle;

[0091] The functions of the gyroscopes and accelerometers are as follows:

[0092] Gyroscope: This is a sensor capable of measuring angular velocity. It can sensitively detect the minute rotations of the device in all directions. For example, when a drone adjusts its attitude in the air, the gyroscope can precisely sense its rotations around various axes. Even the most subtle rotational changes can be captured in a timely manner. It is like a high-precision compass, constantly monitoring the attitude dynamics of the drone.

[0093] Accelerometer: Used to sense linear motion, that is, the translational motion of the device in three-dimensional space. For instance, when the drone is displaced up and down, left and right, or forward and backward due to air flow impact, the accelerometer can quickly detect these position changes. It works in conjunction with the gyroscope, providing comprehensive data support for accurately judging the overall motion state of the drone.

[0094] Data acquisition and transmission are as follows:

[0095] Real-time monitoring: During the flight of the drone, the gyroscope and accelerometer continuously collect data to monitor the attitude changes of the drone in real time. The data acquisition frequency is very high to ensure that any sudden attitude changes can be captured promptly. For example, the data may be collected dozens or even hundreds of times per second to accurately reflect the attitude changes of the drone.

[0096] Data transmission: The collected data is immediately transmitted to the processor of the anti-shake unit through the internal high-speed data bus. This data transmission process requires high speed and stability to ensure that the processor can obtain the latest and accurate attitude information of the drone in the shortest time, providing a basis for subsequent processing and decision-making.

[0097] Data processing and analysis are as follows:

[0098] Data fusion: After the processor receives the data transmitted by the gyroscope and accelerometer, it will first perform data fusion processing on these data. Since the data types and characteristics collected by the two sensors are different, through data fusion algorithms, their advantages can be combined to obtain more accurate and comprehensive attitude information of the drone. For example, when the drone is rotating rapidly, the data of the gyroscope may better reflect its instantaneous attitude changes, while during steady flight, the data of the accelerometer can provide more reliable position information. The data fusion algorithm can reasonably utilize these two types of data according to different flight states.

[0099] Calculate attitude deviation: Based on the fused data, the processor calculates the deviation between the current attitude of the drone and the preset stable attitude. This deviation includes multiple aspects such as angle deviation and displacement deviation. Through the precise calculation of these deviations, specific numerical bases are provided for subsequent compensation adjustments.

[0100] The compensation adjustment mechanism is as follows:

[0101] Generate adjustment instructions: Based on the calculated attitude deviation, the processor generates corresponding adjustment instructions to determine how to compensate and adjust the camera to restore it to a stable locked state. These adjustment instructions include specific parameters such as the direction and amplitude of the adjustment. For example, if the drone tilts to the left, causing the camera to deviate from the target obstacle, the processor calculates the specific angle and displacement required for a right adjustment.

[0102] Drive the actuator: The adjustment instructions are sent to the drive actuator of the camera, such as a motor or an electronic drive device. These actuators precisely control the movement of the camera according to the instructions, adjusting it in the predetermined direction and amplitude. For example, by controlling the rotation of the motor, the lens or sensor of the camera is driven to translate or rotate to compensate for the attitude change of the drone.

[0103] Real-time feedback and optimization are as follows:

[0104] Effect monitoring: During the adjustment of the camera, the anti-shake unit continuously monitors the adjustment effect. By comparing the attitude deviation data before and after the adjustment, it evaluates whether the camera has accurately locked the target obstacle. If the adjustment effect is not ideal, the system further optimizes the adjustment strategy and continues to fine-tune the camera until a satisfactory locked effect is achieved.

[0105] Adaptive adjustment: As the flight environment of the drone changes, the external interference it experiences also changes continuously. The anti-shake unit can adaptively adjust the anti-shake strategy and parameters based on the real-time monitored data to adapt to different flight conditions. For example, in an environment with strong wind, the anti-shake intensity and response speed are increased; during stable flight, the anti-shake intensity is appropriately reduced to improve the energy efficiency and stability of the system.

[0106] In summary, the anti-shake unit continuously monitors the attitude change of the drone through high-precision sensors, drives the actuator for compensation adjustment after data processing and analysis, and has the ability of real-time feedback and optimization, thus ensuring that the camera always stably locks the target obstacle in the complex and changeable flight environment;

[0107] The intelligent algorithm unit is embedded with an intelligent tracking algorithm and combines the data provided by the vision recognition module to enable real-time tracking of the obstacle position. During the flight of the drone, the obstacle may move due to wind force, its own movement, etc. To maintain precise alignment with the obstacle, the intelligent tracking algorithm will adjust the locking position in real time according to the obstacle position information provided by the vision recognition module. After obtaining the obstacle position information, the intelligent tracking algorithm starts to function. This algorithm will calculate the angle that the laser emitter needs to adjust based on the current flight attitude, speed of the drone, and the position information of the obstacle to ensure that the laser beam can accurately point to the target obstacle. Then, by adjusting the flight attitude of the drone, the laser beam can always maintain precise alignment with the obstacle. This dynamic adjustment ability ensures that the laser beam can always accurately act on the target obstacle, and even if the obstacle is moving, a stable locking effect can be maintained;

[0108] In summary, the anti-shake locking module and its multiple units together constitute the key part of the drone laser obstacle clearing system. By means of real-time monitoring, dynamic adjustment, and physical support, etc., it ensures the stable locking of the obstacle by the camera in a complex environment, providing accurate target information for subsequent laser obstacle clearing operations. This efficient and stable anti-shake locking function greatly improves the overall performance and reliability of the drone laser obstacle clearing system.

[0109] Optionally, the ground control terminal includes:

[0110] A real-time monitoring interface unit for displaying the real-time video stream, flight parameters, obstacle recognition results, and laser obstacle clearing progress of the drone;

[0111] A remote control unit for sending control commands to the vision recognition module to remotely command the takeoff, landing, and route planning of the drone, and also for sending obstacle clearing commands to the drone laser obstacle clearing module to perform laser obstacle clearing operations.

[0112] A data analysis and reporting unit responsible for automatically recording the historical data of the obstacle clearing task and generating reports for subsequent analysis and optimization.

[0113] In this optional embodiment, the ground control terminal integrates a real-time monitoring interface unit, a remote control unit, and a data analysis and reporting unit, which is responsible for real-time monitoring of the drone's state, remotely controlling the drone to perform obstacle clearing tasks, and collecting and analyzing task data to optimize subsequent operations. Through the joint cooperation of multiple units, it ensures the efficient and safe operation of the drone laser obstacle clearing system. The specific work content is as follows:

[0114] The real-time monitoring interface unit is used for the intuitive display of the UAV status. This interface can display the video stream of the UAV in real time, enabling the operator to clearly see the transmission line and its surrounding environment from the UAV's perspective. At the same time, the flight parameters of the UAV, including key information such as altitude, speed, and position, are also shown on the interface, helping the operator accurately grasp the flight status of the UAV. In addition, the real-time monitoring interface also shows the obstacle recognition results and the progress of laser obstacle clearance, enabling the operator to timely understand the progress of the obstacle clearance task;

[0115] The remote control unit combines input devices such as touch screens or keyboards and mice, enabling the operator to remotely command the takeoff, landing, flight path planning, and laser obstacle clearance operations of the UAV. The convenience and flexibility of the operation are fully considered, enabling the operator to remotely control the UAV at any time and any place. At the same time, the remote control function unit also has safety protection mechanisms, such as permission verification and emergency shutdown, to ensure the safety of the UAV during the control process;

[0116] The data analysis and reporting unit is responsible for collecting and analyzing the historical data of the UAV obstacle clearance task. It can automatically record key information such as obstacle types, positions, and obstacle clearance efficiency, and generate detailed reports for subsequent analysis and optimization. Through data analysis, the operator can deeply understand the execution situation and efficiency of the UAV obstacle clearance task, discover potential problems and improvement points. At the same time, the report can also provide valuable reference information for future obstacle clearance tasks, helping to optimize the task planning and execution strategy;

[0117] In summary, the ground control terminal module and its multiple units cooperate together. Through real-time monitoring, remote control, and data analysis, it ensures the efficient and safe operation of the UAV laser obstacle clearance system, providing a strong guarantee for the stable operation of the power system.

[0118] As Figure 2 shown, this application also provides a UAV laser obstacle clearance method based on a vision method, and the steps are as follows:

[0119] S1: When the system starts, each module is initialized to ensure that they are in a normal working state;

[0120] S2: The ground control terminal displays the system status. After the operator confirms that it is correct, the task is started;

[0121] S3: The vision recognition module starts to capture and process images, recognizes obstacles, and transmits the target positioning information;

[0122] S4: The UAV laser obstacle clearance module selects a suitable laser emitter to work according to the target positioning information, and accurately emits a laser beam to cut the obstacle;

[0123] S5: The anti-shake locking module ensures the stable locking of the camera on obstacles in complex flight environments, providing accurate target information for laser obstacle clearance. Specifically, a multi-degree-of-freedom flexible gimbal introduces flexible damping materials or non-linear spring structures, mimicking the S-shaped multi-segment structure of a bird's cervical vertebra. Each joint is equipped with a micro hydraulic damping chamber (filled with magnetorheological fluid), and each joint is equipped with a piezoelectric fiber sensor (embedded in a silicone layer) to monitor deformation in real time and feedback it to the control unit. A micro pneumatic artificial muscle is used to replace the traditional motor, achieving rapid contraction or extension through air pressure changes. The driving layout specifically arranges 4 groups of artificial muscles in a cross pattern to form an antagonistic structure similar to biological muscles; the robotic arm uses a foldable pneumatic wind-resistant robotic arm, and the structural design is a variable cross-section robotic arm. The robotic arm uses a telescopic carbon fiber rod with a shape memory polymer skin. The skin can contract and expand under current stimulation, changing the pneumatic shape of the robotic arm, similar to the flaps of an aircraft. Vortex generators (micro sawtooth structures) are set at the folding joints to disrupt the separated airflow to reduce turbulence. The active wind-resistant structure is set with a micro jet nozzle integrated at the end of the robotic arm, which sprays high-speed airflow according to the data of the air pressure sensor to actively counteract the side wind moment. The control logic is that when the wind speed sensor detects a gust of wind, the robotic arm automatically extends to the longest state, increasing the lever arm, and the skin contracts to form an airfoil groove to reduce aerodynamic drag. The jet nozzle sprays according to a pre-programmed pattern to generate a reverse moment to balance the body; the gimbal specifically uses a piezoelectric-electromagnetic hybrid energy recovery gimbal. The structural design uses a vibration energy recovery layer. Piezoelectric ceramic arrays are embedded in the gimbal bracket to convert mechanical vibration into electrical energy. Hall electromagnetic coil groups are superimposed to generate electricity by cutting magnetic induction lines with the swing of the gimbal, forming a complement to the piezoelectric effect. The self-powered circuit stores the recovered electrical energy in a supercapacitor and directly powers the gimbal sensors (such as gyroscopes, encoders), reducing the load on the main battery;

[0124] S6: The ground control terminal monitors information such as the status of the UAV and the progress of obstacle clearance in real time and performs remote control as needed;

[0125] S7: After the obstacle clearance task is completed, the UAV returns to the base or a designated location;

[0126] S8: The ground control terminal collects and analyzes the mission data and generates a detailed report for subsequent analysis and optimization.

[0127] During the entire mission execution process, the built-in safety protection mechanisms in each module (such as the automatic shutdown of the laser emitter, the obstacle avoidance function of the UAV, etc.) ensure the safety of the operation. If an abnormal situation or potential risk occurs, the ground control terminal will immediately issue an alarm and take corresponding safety measures.

[0128] The above-mentioned UAV laser obstacle clearing method based on the vision method can implement each embodiment of the above-mentioned UAV laser obstacle clearing system based on the vision method, and can achieve the same beneficial effects, which will not be elaborated here.

[0129] The preferred specific embodiments of the present invention have been described in detail above. It should be understood that those of ordinary skill in the art can make many modifications and variations according to the concept of the present invention without creative work. Therefore, all technical solutions that can be obtained by those skilled in the art in the technical field of the present invention based on the concept of the present invention through logical analysis, reasoning or limited experiments on the basis of the prior art shall fall within the protection scope determined by the claims.

Claims

1. A drone laser obstacle clearing system based on a vision method, characterized in that, Comprising: A visual recognition module, configured to recognize obstacle information and generate target positioning information based on the obstacle information; An unmanned aerial vehicle (UAV) laser obstacle clearing module, configured to emit a laser beam to cut and clear obstacles on a transmission line according to the target positioning information provided by the visual recognition module; An anti-shake locking module, configured to control the visual recognition module to stably lock on to obstacles in a complex flight environment; A ground control terminal, configured to send control instructions to the visual recognition module, and also configured to monitor the state of the UAV in real time, remotely control the UAV to perform the obstacle clearing task, and collect and analyze task data.

2. The drone laser obstacle clearing system based on a vision method according to claim 1, characterized in that The visual recognition module includes a visual camera, a lidar, a spectrometer, and a thermal imaging sensor; The visual camera is configured to, after receiving the control instructions from the ground control terminal, capture, process, and analyze obstacle image information in real time, and is also configured to, based on the obstacle image information and the three-dimensional coordinates of the obstacle calculated by the lidar, identify the material type through a spectrometer and a laser-induced breakdown spectroscopy probe, and through laser reflection spectroscopy; The thermal imaging sensor is configured to identify live wires and humans by fusing RGB, LiDAR, and thermal imaging data through a deep learning model, and optimize the obstacle information and the target positioning information according to the recognition results.

3. The drone laser obstacle clearing system based on a vision method according to claim 2, characterized in that, The visual camera includes a high-definition camera and an image processing unit. The high-definition camera is configured to capture real-time images of the transmission line and the surrounding environment; the image processing unit is configured to perform preprocessing, feature extraction, obstacle recognition, and target positioning on the real-time images to obtain target positioning information.

4. A drone laser obstacle clearing system based on a vision method according to claim 1, characterized in that The obstacle information includes the size and material of the obstacle. The UAV laser obstacle clearing module includes: A multi-laser emission unit, configured to configure laser emitters with different powers and wavelengths according to the size and material of the obstacle; A laser automatic aiming module, configured to aim the multi-laser emission unit at the target according to the target information to guide the emission of the laser beam; A stable flight platform, equipped with a GPS navigation, an inertial navigation system, and an obstacle avoidance sensor, to ensure the stable flight of the UAV in a complex environment.

5. The drone laser obstacle clearing system based on a vision method according to claim 1, characterized in that, The anti-shake locking module includes: A center-of-gravity adjustment and stabilization unit, provided with a movable counterweight, configured to balance the fuselage in real time according to the movable counterweight; An automatic anti-shake unit, built-in with a gyroscope and an accelerometer, configured to monitor the attitude change of the UAV in real time, and automatically adjust the image capture attitude to maintain stable locking on to the obstacle.

6. The drone laser obstacle clearing system based on a vision method according to claim 1, characterized in that, The anti-shake locking module further includes a multi-degree-of-freedom flexible gimbal. The structural type of the multi-degree-of-freedom flexible gimbal is a flexible damping material or a non-linear spring structure, including an S-shaped multi-segment structure. Each adjacent two segments are connected by a joint. Each joint is internally provided with a micro hydraulic damping cavity. Each joint is equipped with a piezoelectric fiber sensor for real-time monitoring of deformation and feedback to the control unit. A micro pneumatic artificial muscle is used, and rapid contraction or extension is achieved through air pressure change. The driving layout specifically uses 4 groups of artificial muscles arranged in a cross to form an antagonistic structure similar to biological muscles. The robotic arm of the multi-degree-of-freedom flexible gimbal uses a foldable pneumatic wind-resistant robotic arm. A vortex generator is set at the folding joint to disrupt the separated air flow to reduce turbulence. The active wind-resistant structure is set that a micro jet nozzle is integrated at the end of the robotic arm, and high-speed air flow is ejected according to the data of the air pressure sensor to actively offset the side wind moment. The multi-degree-of-freedom flexible gimbal specifically uses a piezoelectric-electromagnetic hybrid energy recovery gimbal. The structural design uses a vibration energy recovery layer. A piezoelectric ceramic sheet array is embedded in the gimbal bracket to convert mechanical vibration into electrical energy. A Hall electromagnetic coil group is superimposed, and electricity is generated by using the gimbal swing to cut the magnetic induction line, forming a complement with the piezoelectric effect.

7. A drone laser obstacle clearing system based on a vision method according to claim 1, characterized in that, The ground control terminal includes: A real-time monitoring interface unit for displaying the real-time video stream, flight parameters, obstacle recognition results, and laser obstacle clearing progress of the unmanned aerial vehicle; A remote control unit for sending control commands to the visual recognition module to remotely command the takeoff, landing, and route planning of the unmanned aerial vehicle, and also for sending obstacle clearing commands to the unmanned aerial vehicle laser obstacle clearing module to perform laser obstacle clearing operations.

8. A drone laser obstacle clearing system based on a vision method according to claim 1, characterized in that, The ground control terminal further includes: A data analysis and reporting unit responsible for automatically recording the historical data of the obstacle clearing task and generating reports for subsequent analysis and optimization.

9. A method for clearing obstacles of an unmanned aerial vehicle based on a vision method, characterized in that, Including: S1: When the system starts, each module is initialized to ensure that they are in a normal working state; S2: The ground control terminal displays the system status, and after the operator confirms that it is correct, the task starts to be executed; S3: The visual recognition module starts to capture and process images, identifies obstacles, and transmits target positioning information; S4: The unmanned aerial vehicle laser obstacle clearing module selects a laser emitter to work according to the target positioning information and accurately emits a laser beam to cut the obstacle; S5: The anti-shake locking module ensures the stable locking of the camera on the obstacle in a complex flight environment and provides accurate target information for laser obstacle clearing; S6: The ground control terminal monitors the status, obstacle clearing progress and other information of the unmanned aerial vehicle in real time and performs remote control as needed; S7: When the obstacle clearing task is completed, the unmanned aerial vehicle returns to the base or a designated location; S8: The ground control terminal collects and analyzes the task data and generates a detailed report for subsequent analysis and optimization.

Images