Forest stereo survey device and method based on double unmanned aerial vehicle cooperation

Abstract

This invention belongs to the field of forest survey technology, and particularly relates to a forest three-dimensional survey device and method based on dual-UAV collaboration. The understory UAV transmits real-time video streams of the forest understory environment, spectral images of understory vegetation, and three-dimensional point clouds of the understory to the above-story UAV. The above-story UAV acquires its own spatial information, the three-dimensional structure of the forest canopy, and the spectral information of the forest canopy. It then transmits the information collected by the understory UAV and its own information to a ground control base station. The ground control base station coordinates the dual-UAV platforms to obtain a real-world forest model. This invention, through a collaborative mode of two-way communication and dynamic following between the two UAVs, ensures that the understory UAVs can safely, continuously, and efficiently conduct survey operations in complex environments.

Description

Technical Field

[0001] This invention belongs to the field of forest survey technology, and in particular relates to a forest three-dimensional survey device and method based on dual unmanned aerial vehicle (UAV) collaboration. Background Technology

[0002] Traditional forest survey methods, such as ground-based artificial plot surveys, suffer from low efficiency and high risk. Satellite remote sensing or single-UAV aerial photography are affected by canopy shading, making it difficult to effectively obtain key information such as forest understory structure, seedling regeneration, and vertical biomass distribution. Existing UAV-based forest survey schemes mainly face two major technical bottlenecks: first, single UAVs face multiple challenges such as lost positioning signals, communication interruptions, and difficulty in ensuring flight safety when flying in complex understory environments; second, the observed data from the forest canopy and understory are fragmented, making it difficult to achieve a complete, three-dimensional digital perception of the forest ecosystem.

[0003] To address the aforementioned issues, this proposal presents a forest survey device and method based on dual-UAV communication and collaboration, assuming that "over-forest UAVs can acquire stable satellite signals and wide-area communication capabilities, while under-forest UAVs need to complete detailed surveys in satellite-denied environments." The core concept of this scheme is to construct an aerial collaborative system where over-forest UAVs act as "communication relay stations" and "canopy observation stations," while under-forest UAVs act as "under-forest explorers." Through two-way communication and dynamic tracking, the system ensures the safe, continuous, and efficient operation of under-forest UAVs, ultimately achieving synchronous acquisition and integrated modeling of forest canopy and understory structure data. Summary of the Invention

[0004] In view of this, the present invention aims to provide a forest three-dimensional survey device and method based on dual-UAV collaboration to address the shortcomings of existing UAV-based forest survey schemes. Based on the practical application premise that UAVs above the forest need stable satellite signals and wide-area communication capabilities, and that UAVs below the forest need to complete detailed exploration in satellite-denied environments, the present invention constructs an aerial collaborative system. The UAVs above the forest serve as communication relay stations and canopy observation stations, while the UAVs below the forest serve as detailed exploration terminals. Through a collaborative mode of two-way communication and dynamic following between the two UAVs, the system ensures that the UAVs below the forest canopy can safely, continuously, and efficiently conduct exploration operations in complex environments. Ultimately, this achieves synchronous acquisition of forest canopy and understory structure data, as well as integrated digital modeling of the forest ecosystem.

[0005] To achieve the above objectives, the technical solution created by this invention is implemented as follows: A forest three-dimensional survey device based on dual-UAV collaboration includes a UAV in the forest, a UAV under the forest, and a ground control base station. The UAV in the forest establishes two-way communication with both the UAV under the forest and the ground control base station. The understory drone is used to acquire and transmit real-time video streams of the understory environment, spectral images of understory vegetation, and 3D point clouds of the understory to the above-story drone. The above-story drone is used to acquire its own spatial information, the 3D structure of the forest canopy, and the spectral information of the forest canopy. It then transmits these data to the ground control base station. The ground control base station monitors and visualizes the working status of both the understory and above-story drones in real time and sends control commands to the above-story drone. The above-story drone forwards the control commands and its spatial information to the understory drone, enabling the ground control base station to coordinate the control of the two drone platforms and obtain a real-world forest model.

[0006] Furthermore, the understory drone includes a first lidar, a first antenna array, a first multispectral camera, and a first airborne processing unit. The first lidar is used to perform real-time three-dimensional scanning of the understory environment to obtain three-dimensional point cloud data, and to construct and update the understory three-dimensional point cloud in real time based on the three-dimensional point cloud data. The first multispectral camera is used to collect understory vegetation spectral images and understory image data in real time according to a preset mode. The first airborne processing unit is used to store understory vegetation spectral images, understory image data, and understory three-dimensional point clouds in real time, and to generate a real-time video stream of the understory environment based on the understory image data. The first antenna array is used to construct a data link between the understory drone and the overstory drone, and to transmit the real-time video stream of the understory environment, understory vegetation spectral images, and understory three-dimensional point clouds to the overstory drone.

[0007] Furthermore, the forest-based UAV includes a second lidar, a second second antenna array, a second multispectral camera, a second airborne processing unit, a GNSS module, and a 4G / 5G communication antenna. The second lidar is used to perform real-time three-dimensional scanning of the forest canopy and construct the three-dimensional structure of the canopy based on the scanning results. The second multispectral camera is used to acquire the spectral information of the forest canopy. The second airborne processing unit is used to store the three-dimensional structure and spectral information of the forest canopy in real time and transmit them to the 4G / 5G communication antenna. The GNSS module is used to acquire the spatial information of the forest-based UAV in real time and transmit it to the 4G / 5G communication antenna. The 5G communication antenna and the second antenna array are used to receive real-time video streams of the forest environment, spectral images of forest vegetation, and 3D point clouds of the forest. The received information is forwarded to the 4G / 5G communication antenna. The 4G / 5G communication antenna is used to transmit the spatial information of the drone in the forest, the real-time video streams of the forest environment, the spectral images of forest vegetation, the 3D point clouds of the forest, the 3D structure of the forest canopy, and the spectral information of the forest canopy to the ground control base station. It also receives control commands from the ground control base station and sends the control commands to the second antenna array to control the drone in the forest. The second antenna array also sends the control commands and the real-time spatial information of the drone in the forest to the first antenna array to control the drone in the forest.

[0008] Furthermore, the ground control base station includes a communication ground station and a graphics workstation. The communication ground station is used to receive spatial information of the UAVs above the forest, real-time video streams of the forest understory environment, spectral images of forest understory vegetation and 3D point clouds of the forest understory, 3D structure of the forest canopy and spectral information of the forest canopy. It also transmits control commands issued by the operator through the control software to the UAVs above the forest, which in turn forward them to the UAVs below the forest, thus achieving coordinated control of the two UAV platforms. The graphics workstation is used to receive spatial information of the UAVs above the forest, real-time video streams of the forest understory environment, spectral images of forest understory vegetation and 3D point clouds of the forest understory, 3D structure of the forest canopy and spectral information of the forest canopy forwarded by the communication ground station. Based on the received information, it monitors the working status of the UAVs above and below the forest in real time, and performs data fusion and modeling on the received information to generate and visualize a real-world forest model.

[0009] A method for three-dimensional forest survey based on dual-UAV collaboration, utilizing a three-dimensional forest survey device based on dual-UAV collaboration, specifically includes the following steps: S1: The operator plans the target path for the understory drone on the ground control base station and makes the overstory drone automatically follow the movement of the understory drone. The first flight position of the overstory drone is directly above the canopy corresponding to the first path point of the understory drone. S2: The operator sends control commands to the UAV above the forest via the ground control base station. The UAV above the forest forwards the control commands to the UAV below the aircraft, and the UAV below the aircraft flies autonomously according to the target path. S3: The understory drone collects real-time video streams of the understory environment, spectral images of understory vegetation, and three-dimensional point clouds of the understory at time t, and transmits the collected information at time t back to the ground control base station via the understory drone. S4: Based on the comprehensive channel quality of the drones above and below the forest at time t, adjust the relative positions of the drones above and below the forest in real time to ensure optimal channel quality for both drones above and below the forest. S5: At time t, the UAV in the forest collects its own spatial information, the three-dimensional structure of the forest canopy, and the spectral information of the forest canopy, and transmits the collected information at time t back to the ground control base station. The ground control station performs data fusion and modeling on the collected results of the UAV under the forest and the UAV in the forest according to the time sequence to generate a real forest scene model. S6: Replace time t with time t+1, and repeat steps S3-S5 until the drone below the aircraft completes its flight mission along the target path. The operator controls the drone below the aircraft to return to its home base via the ground control base station. At the same time, the drone below the aircraft guides the drone on the aircraft back to its respective take-off and landing point.

[0010] Furthermore, step S4 specifically includes the following steps: S41: Let the horizontal position of the UAV in the forest at time t be... The horizontal position of the drone under the forest at time t is The horizontal distance between the drones above and below the forest is The overall channel quality at time t is Q(t); S42: Calculate the expected communication distance at time t based on the comprehensive channel quality at time t: =clip( ( ), , ); in, Let be the expected communication distance at time t. To preset the optimal channel quality, Sensitivity coefficient and These represent the lower limit of the safe distance and the upper limit of the effective communication distance, respectively. `clip(x,a,b)` means restricting the variable `x` to the interval [a,b], where `a` and `b` are the upper and lower limits of the interval, respectively. To preset the theoretically optimal communication distance; S43: Construct the joint cost function and solve for the optimal adjustment using gradient descent: ; ; ; ; in, To achieve the optimal adjustment amount, For the joint cost function, For distance error term, This is the channel quality prediction error term. For dynamic weighting coefficients, Adjust the vector for position. The regularization coefficient is . ( This is a local channel quality spatial prediction model based on historical and real-time detection data. To preset the optimal channel quality; S44: Based on the optimal adjustment amount, the target position of the drone above the forest is calculated using the following formula, and the drone above the forest is driven to fly to the target position based on the calculation result, thus completing the adjustment of the relative position between the drone above the forest and the drone below the forest: ; in, The target location for the drone in the forest.

[0011] Compared with the prior art, the present invention can achieve the following beneficial effects: (1) The forest three-dimensional survey device and method based on dual UAV collaboration created by the present invention fundamentally solves the systemic problems of long-standing issues in forest resource surveys, such as fragmented observation dimensions, difficulties in complex environments, and low overall efficiency, compared with the prior art.

[0012] (2) The forest three-dimensional survey device and method based on dual UAV collaboration described in this invention achieves full-dimensional, synchronized, and high-precision integrated data acquisition from the top of the forest canopy to the surface topography. This breakthrough stems from the establishment of a "dual-machine collaborative three-dimensional observation architecture". In the prior art, satellite remote sensing, single UAV aerial photography, or ground manual surveys cannot simultaneously acquire continuous data of the complete vertical profile of the forest due to the single observation perspective and technical means, resulting in a long-term data separation between canopy information and understory structure. However, this invention achieves the simultaneous development of canopy scanning and understory detection tasks by deploying physically separate but logically deeply collaborative dual observation units: the UAV in the forest provides a stable absolute spatial reference in open airspace using a high-precision GNSS receiver; the UAV in the understory is equipped with a lidar and a multispectral imaging module to collect refined three-dimensional point cloud and spectral information in complex and obscured environments. The two types of observation data streams are precisely correlated in time and space through a dedicated data link, and after integrated fusion processing at the back end, a seamless forest full-element real-scene model is generated. This architecture fundamentally breaks through the dimensional limitations of traditional observation methods, providing an unprecedented and complete data foundation for forest carbon sequestration, accurate biomass estimation, and in-depth biodiversity research.

[0013] (3) The forest three-dimensional survey device and method based on dual UAV collaboration described in this invention ensures the operational safety and mission continuity of UAVs under the forest canopy in environments where satellite signals are denied. This key advantage is mainly due to the invention of the "dynamic communication relay and adaptive cooperative position-keeping mechanism". Under dense forest canopies, traditional UAVs are prone to loss of positioning signals and interruption of communication links, which greatly limits their operational depth and flight safety. This invention uses the UAVs in the forest as dynamic aerial relay stations to establish a stable dual-hop communication link. More importantly, the system can effectively offset the signal fading problem caused by branches and leaves through continuous channel quality estimation and dynamic adaptation of communication parameters. At the same time, relying on the precise distance measurement technology between the two UAVs, the relative positions of the two UAVs are monitored in real time. When the distance exceeds the threshold, the UAV in the forest will be automatically triggered to follow the movement, ensuring that the UAV under the forest is always in the optimal communication coverage range. This closed-loop control mechanism provides an uninterrupted command and communication link and a rough external positioning reference for forest operations, enabling forest drones to safely and autonomously complete long-term, large-scale exploration tasks even in environments with no satellite signals. This fundamentally overcomes the core technical obstacle of single drones operating deep in the forest.

[0014] (4) The forest three-dimensional survey device and method based on dual UAV collaboration described in this invention significantly improves the automation level and overall efficiency of the survey operation, mainly due to the meticulous design of the "systematic parallel collaborative workflow". Traditional survey methods often require multiple, phased operations, resulting in high manpower and time costs. This invention integrates dual-drone takeoff, synchronous data acquisition, dynamic collaboration, and mission return into a highly automated parallel process. Through the integrated task planning and execution mode, not only is the efficiency of field work increased several times, but the dependence on continuous manual operation by professional drone pilots is also reduced, significantly lowering the cost and threshold for conducting large-scale, high-precision forest surveys. Attached Figure Description

[0015] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments and descriptions of the invention are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings: Figure 1 A schematic diagram of the structure of the forest three-dimensional survey device based on dual UAV collaboration as described in an embodiment of the present invention; Figure 2 This is a schematic diagram of the structure of the forest three-dimensional survey method based on dual UAV collaboration described in an embodiment of the present invention.

[0016] Explanation of reference numerals in the attached figures: 1. Understory UAV; 2. Abovestory UAV; 3. Ground control base station; 11. First lidar; 12. First antenna array; 13. First multispectral camera; 14. First airborne processing unit; 21. Second lidar; 22. Second airborne processing unit; 23. Second multispectral camera; 24. GNSS module; 25. Second antenna array; 26. 4G / 5G communication antenna; 31. Communication ground station; 32. Graphics workstation. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and do not constitute a limitation thereof.

[0018] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other.

[0019] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation on this invention. Furthermore, the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, features defined with "first," "second," etc., may explicitly or implicitly include one or more of that feature. In the description of this invention, unless otherwise stated, "a plurality of" means two or more.

[0020] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art will understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0021] The invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0022] like Figure 1 As shown, this invention provides a forest three-dimensional survey device based on dual-UAV collaboration, including a forest-based UAV 2, a forest-based UAV 1, and a ground control base station 3. The forest-based UAV 2 establishes two-way communication with both the forest-based UAV 1 and the ground control base station 3. Understory drone 1 is used to acquire and transmit real-time video streams of the understory environment, spectral images of understory vegetation, and 3D point clouds of the understory to forest-top drone 2. Forest-top drone 2 is used to acquire its own spatial information, the 3D structure of the forest canopy, and the spectral information of the forest canopy. It also transmits the real-time video streams of the understory environment, spectral images of understory vegetation, 3D point clouds of the understory, the spatial information of forest-top drone 2, the 3D structure of the forest canopy, and the spectral information of the forest canopy to ground control base station 3. Ground control base station 3 is used to monitor and visualize the working status of understory drone 1 and forest-top drone 2 in real time, and sends control commands to forest-top drone 2. Forest-top drone 2 forwards the control commands and its spatial information to understory drone 1, realizing the collaborative control of the two drone platforms by ground control base station 3 and obtaining a real-world forest model.

[0023] It should be noted that the forest three-dimensional survey device based on dual-UAV collaboration includes a forest canopy UAV 1 (specifically, a forest canopy UAV 1 detection platform), a forest relay UAV 1 (specifically, a forest-above relay UAV platform), and a ground control base station 3. The forest canopy UAV 1 utilizes a first lidar 11 to perform forest canopy scanning and modeling, which is also used for path planning. It is equipped with a first multispectral camera 13 to scan the forest environment and communicates with the forest-above UAV 2 via a first antenna array 12 for real-time video transmission. The forest-above UAV 2 is equipped with a second antenna array 25 and a 4G / 5G communication antenna 26 to relay signals between the forest canopy UAV 1 and the ground control base station 3. It is also equipped with a second lidar 21 to perform forest canopy scanning and a second multispectral camera 23 to perform canopy spectral measurements.

[0024] In some embodiments, the forest drone 1 includes a first lidar 11, a first antenna array 12, a first multispectral camera 13, and a first airborne processing unit 14. The first lidar 11 is used to perform real-time three-dimensional scanning of the forest environment to obtain three-dimensional point cloud data, and to construct and update the forest three-dimensional point cloud in real time based on the three-dimensional point cloud data. The first multispectral camera 13 is used to collect forest vegetation spectral images and forest image data in real time according to a preset mode. The first airborne processing unit 14 is used to store forest vegetation spectral images, forest image data, and forest three-dimensional point clouds in real time, and to generate a real-time video stream of the forest environment based on the forest image data. The first antenna array 12 is used to construct a data link between the forest drone 1 and the forest drone 2, and to transmit the real-time video stream of the forest environment, forest vegetation spectral images, and forest three-dimensional point clouds to the forest drone 2.

[0025] It should be noted that the understory drone 1 is the main component for in-depth, detailed exploration within the forest. It is equipped with a first lidar 11 and a first multispectral camera 13, used to acquire high-precision 3D point cloud and spectral images of understory vegetation in real time, respectively. The first lidar 11 provides 3D environmental modeling, guiding the understory drone 1 to achieve autonomous positioning and stable flight. A directional multi-antenna array is responsible for establishing a dedicated data link with the above-the-forest drone 2. The embedded first airborne processing unit 14 processes the point cloud data in real time to achieve instant positioning, mapping, and autonomous obstacle avoidance, while also storing the data. The real-time video stream acquired by the first multispectral camera 13 can be transmitted uplink via the dedicated data link. The adaptive RF communication module of the multi-antenna array is designed based on the AD9361 chip, supporting 700MHz to 6GHz wideband dynamic frequency hopping and possessing spectrum sensing and frequency hopping mechanisms.

[0026] In some embodiments, the forest drone 2 includes a second lidar 21, a second second antenna array 25, a second multispectral camera 23, a second airborne processing unit 22, a GNSS module 24, and a 4G / 5G communication antenna 26. The second lidar 21 is used to perform real-time three-dimensional scanning of the forest canopy and construct the three-dimensional structure of the forest canopy in real time based on the scanning results. The second multispectral camera 23 is used to acquire the spectral information of the forest canopy. The second airborne processing unit 22 is used to store the three-dimensional structure and spectral information of the forest canopy in real time and transmit the three-dimensional structure and spectral information of the forest canopy to the 4G / 5G communication antenna 26. The GNSS module 24 is used to acquire the spatial information of the forest drone 2 in real time and transmit the spatial information of the forest drone 2. The second antenna array 25 is used to receive real-time video streams of the forest environment, spectral images of forest vegetation, and three-dimensional point clouds of the forest, and forwards the received information to the 4G / 5G communication antenna 26. The 4G / 5G communication antenna 26 is used to transmit the spatial information of the forest drone 2, the real-time video streams of the forest environment, the spectral images of forest vegetation, the three-dimensional point clouds of the forest, the three-dimensional structure of the forest canopy, and the spectral information of the forest canopy to the ground control base station 3. It also receives control commands issued by the ground control base station 3 and sends the control commands to the second antenna array 25 to control the forest drone 2. The second antenna array 25 also sends the control commands and the real-time spatial information of the forest drone 2 to the first antenna array 12 to control the forest drone 1.

[0027] It should be noted that the forest drone 2 serves as an aerial communication hub and canopy observation station. Its hardware configuration revolves around two core functions: first, a communication relay function, achieved through two independent communication systems—a high-performance multi-antenna array for establishing a dedicated network link with the forest drone 1, and a long-range 4G / 5G communication antenna 26 or a bridge antenna for maintaining a wide-area connection with the ground control base station 3. Second, a canopy observation function, simultaneously equipped with a second lidar 21 and a second multispectral camera 23 to acquire the three-dimensional structure and spectral information of the upper forest canopy. The forest drone 2 platform also integrates a high-precision GNSS module 24, providing a precise absolute coordinate reference for the entire collaborative system.

[0028] In some embodiments, the ground control base station 3 includes a communication ground station 31 and a graphics workstation 32. The communication ground station 31 is used to receive spatial information of the forest drone 2, real-time video stream of the forest environment, spectral images of forest vegetation and 3D point clouds of the forest, 3D structure of the forest canopy and spectral information of the forest canopy, and transmit the control commands issued by the operator through the control software to the forest drone 2, which is then forwarded by the forest drone 2 to the forest drone 1, thereby realizing the collaborative control of the two drone platforms. The graphics workstation 32 is used to receive the spatial information of the forest drone 2, real-time video stream of the forest environment, spectral images of forest vegetation and 3D point clouds of the forest, 3D structure of the forest canopy and spectral information of the forest canopy forwarded by the communication ground station 31, and to monitor the working status of the forest drone 1 and the forest drone 2 in real time based on the received information, and to perform data fusion and modeling on the received information to generate and visualize a real forest scene model.

[0029] It should be noted that ground control base station 3 is the command and data processing center of the entire system. In terms of hardware, it mainly consists of a dual-channel communication ground station 31 and a high-performance graphics workstation 32. It runs integrated mission control software, monitors the status of the two UAVs, displays real-time video and trajectory, and provides functions such as one-click mission planning and manual control.

[0030] like Figure 2 As shown, the present invention also provides a method for three-dimensional forest survey based on dual-UAV collaboration, which is implemented using a three-dimensional forest survey device based on dual-UAV collaboration, and specifically includes the following steps: S1: The operator plans the target path for the understory drone 1 on the ground control base station 3, and makes the forest drone 2 automatically follow the movement of the understory drone 1. The first flight position of the forest drone 2 is directly above the canopy corresponding to the first path point of the understory drone 1. S2: The operator sends control commands to the forest drone 2 through the ground control base station 3. The forest drone 2 forwards the control commands to the drone below it, and the drone below it flies autonomously according to the target path. S3: The understory drone 1 collects real-time video streams of the understory environment, spectral images of understory vegetation, and three-dimensional point clouds of understory at time t, and transmits the collected information at time t back to the ground control base station 3 via the forest drone 2. S4: Based on the comprehensive channel quality of UAV 2 above the forest and UAV 1 below the forest at time t, adjust the relative positions of UAV 2 above the forest and UAV 1 below the forest in real time to ensure that the channel quality of UAV 2 above the forest and UAV 1 below the forest is optimal. S5: At time t, the forest drone 2 collects its own spatial information, the three-dimensional structure of the forest canopy, and the spectral information of the forest canopy, and transmits the collected information at time t back to the ground control base station 3. The ground control station performs data fusion and modeling on the collected results of the understory drone 1 and the collected results of the forest drone 2 according to the time sequence to generate a real forest scene model. S6: Replace time t with time t+1, and repeat steps S3-S5 until the unmanned aerial vehicle (UAV) on the ground completes its flight mission along the target path. The operator then controls the unmanned aerial vehicle to return to its home base via ground control base station 3. At the same time, the unmanned aerial vehicle guides the UAV on the aircraft back to its respective take-off and landing point.

[0031] It should be noted that the main operational steps of the forest survey method are: (1) mission planning and system initialization; (2) dual-aircraft sequential take-off and link establishment; (3) autonomous understory detection and data collection; (4) dynamic communication support and collaborative flight; (5) synchronous scanning of the canopy; (6) mission termination and collaborative return; (7) data fusion and modeling.

[0032] Specifically, the steps are as follows: Step 1: Task Planning and System Initialization Before the operation, the operator completes the collaborative task planning in the corresponding software of the ground control base station 3. The operator plans the preset path and target area for the understory drone 1, obtaining the target path. The overstory drone 2 automatically follows the flight path of the understory drone 1 by default, with its default first flight position being above the canopy of the first path point of the understory drone 1. After planning is completed, the system performs a dual-drone self-check to ensure that the communication link is unobstructed and all sensors are functioning normally.

[0033] Step 2: Sequential takeoff of the two aircraft and establishment of the communication link The system employs a staggered takeoff strategy to ensure safety. Ground control base station 3 first instructs the forest drone 2 to take off, hover at a predetermined altitude above the target forest area, and establish stable wide-area communication between the forest drone 2 and the ground. Subsequently, it instructs the understory drone 1 to take off to a low altitude of 1.5-2 meters above the ground, automatically activates lidar obstacle avoidance and positioning, and establishes a dedicated data link with the already positioned forest drone 2.

[0034] Step 3: Autonomous Detection and Data Collection in the Forest Understory The unmanned aerial vehicle (UAV) 1 flies autonomously along the planned target path. During flight, the first lidar 11 continuously performs 3D scanning of the environment, constructing and updating the 3D point cloud of the surrounding environment in real time for navigation and obstacle avoidance; the first multispectral camera 13 acquires image data according to a predetermined mode, which is timed acquisition at a fixed frequency. While all raw data is stored in the first airborne processing unit 14, key status information and low-latency real-time video streams are relayed back to the ground control base station 3 in real time via a communication link.

[0035] Step 4: Dynamic Communication Support and Coordinated Flight This step is crucial for maintaining system stability. Continuous bidirectional channel quality assessment is performed on both drones, and the communication frequency and modulation scheme are dynamically adjusted accordingly to counteract signal attenuation caused by the forest canopy. Simultaneously, drone 1 (under the trees) monitors the signal strength (SINR) and relative distance between itself and drone 2 (above the trees) in real time. Based on the signal strength and horizontal distance between the two drones, the follow-fly strategy is automatically adjusted, guiding drone 2 to move horizontally to the area directly above drone 1, ensuring the relay link remains optimal at all times.

[0036] Step 5: Simultaneous scanning of the forest canopy During the operation of the understory drone 1, the forest drone 2, guided by the understory drone, simultaneously performs canopy lidar scanning and multispectral imaging tasks.

[0037] Step Six: Mission Termination and Coordinated Return Upon completion of the mission, ground control base station 3 sends a return command. The understory UAV 1 can autonomously navigate to the nearest forest gap or open area and return independently after reacquiring satellite signals; or, while maintaining a relative position with the overstory UAV 2, it can be guided and coordinated by the latter for a return. After confirming the safety of the understory UAV 1, the overstory UAV 2 autonomously returns to its take-off and landing point.

[0038] Step 7: Data Fusion and Modeling After all data is transmitted back to ground control base station 3, the point cloud and image data acquired by the above-forest UAV 2 and the below-forest UAV 1 are precisely registered and fused using a unified timestamp and spatial coordinate reference. Through professional software processing, a seamless forest landscape model from the top of the canopy to the ground surface is finally generated, and multispectral information can be integrated for multidimensional analysis.

[0039] In some embodiments, step S4 specifically includes the following steps: S41: Let the horizontal position of the forest drone 2 at time t be... The horizontal position of the understory drone 1 at time t is The horizontal distance between the above-ground drone 2 and the below-ground drone 1 is The overall channel quality at time t is Q(t); S42: Calculate the expected communication distance at time t based on the comprehensive channel quality at time t: =clip( ( ), , ); in, Let be the expected communication distance at time t. To preset the optimal channel quality, Sensitivity coefficient and These represent the lower limit of the safe distance and the upper limit of the effective communication distance, respectively. `clip(x,a,b)` means restricting the variable `x` to the interval [a,b], where `a` and `b` are the upper and lower limits of the interval, respectively. To preset the theoretically optimal communication distance; S43: Construct the joint cost function and solve for the optimal adjustment using gradient descent: ; ; ; ; in, To achieve the optimal adjustment amount, For the joint cost function, For distance error term, This is the channel quality prediction error term. For dynamic weighting coefficients, Adjust the vector for position. The regularization coefficient is . ( This is a local channel quality spatial prediction model based on historical and real-time detection data. To preset the optimal channel quality; S44: Based on the optimal adjustment amount, the target position of the forest drone 2 is calculated using the following formula, and the forest drone 2 is driven to fly to the target position based on the calculation result, thus completing the adjustment of the relative position between the forest drone 2 and the under-forest drone 1: ; in, The target location of the forest drone 2.

[0040] It should be noted that the horizontal position of the forest drone 2 at time t is... The horizontal position of drone 1 under the forest is The horizontal distance between the two machines is The real-time integrated channel quality index is Q(t).

[0041] First, the expected communication distance is dynamically calculated based on the real-time channel quality Q(t). : =clip( ( ), , ) in, To preset the theoretically optimal communication distance, To preset the optimal channel quality, This is an adjustable sensitivity coefficient. and These represent the lower limit of the safe distance and the upper limit of the effective communication distance, respectively. The function clip(x,a,b) restricts the variable x to the interval [a,b].

[0042] Define the position adjustment vector as Construct the following joint cost function To find the optimal adjustment amount: ； in, This is the distance error term. This is the channel quality prediction error term. ( This is a spatial prediction model for local channel quality based on historical and real-time detection data. λ is a dynamic weighting coefficient whose value is dynamically adjusted according to Q(t): when Q(t) is lower than the first threshold, λ approaches 0, and the system prioritizes optimizing channel quality; when Q(t) is higher than the second threshold, λ approaches 1, and the system prioritizes maintaining distance tracking. λ is the regularization coefficient used to smooth the control output. Specifically, a Q(t)-λ lookup table is established to adjust λ.

[0043] The cost function is solved online using the gradient descent method. Minimize the optimal adjustment amount : ; The target location of the UAV 2 on the forest is: ; The flight control system drives the forest drone 2 to the target location.

[0044] It should be understood that the various forms of processes shown above can be used to reorder, add, or delete steps. For example, the steps described in this invention disclosure can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this invention can be achieved, and this is not limited herein.

[0045] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A forest three-dimensional survey device based on dual-UAV collaboration, characterized in that: This includes aerial drones, understory drones, and ground control base stations. The aerial drones establish two-way communication with both the understory drones and the ground control base stations. The understory drone is used to acquire and transmit real-time video streams of the understory environment, spectral images of understory vegetation, and 3D point clouds of the understory to the above-story drone. The above-story drone is used to acquire its own spatial information, the 3D structure of the forest canopy, and the spectral information of the forest canopy. It then transmits these data to the ground control base station. The ground control base station monitors and visualizes the working status of both the understory and above-story drones in real time and sends control commands to the above-story drone. The above-story drone forwards the control commands and its spatial information to the understory drone, enabling the ground control base station to coordinate the control of the two drone platforms and obtain a real-world forest model.

2. The forest three-dimensional survey device based on dual UAV collaboration according to claim 1, characterized in that: The forest-based drone includes a first lidar, a first antenna array, a first multispectral camera, and a first airborne processing unit. The first lidar is used to perform real-time 3D scanning of the forest environment to obtain 3D point cloud data, and to construct and update the forest 3D point cloud in real time based on the 3D point cloud data. The first multispectral camera is used to collect spectral images of forest vegetation and forest image data in real time according to a preset mode. The first airborne processing unit is used to store spectral images of forest vegetation, forest image data, and forest 3D point cloud in real time, and to generate a real-time video stream of the forest environment based on the forest image data. The first antenna array is used to establish a data link between the forest-based drone and the drone above the forest, and to transmit the real-time video stream of the forest environment, spectral images of forest vegetation, and forest 3D point cloud to the drone above the forest.

3. The forest three-dimensional survey device based on dual UAV collaboration according to claim 2, characterized in that: The forest-based UAV includes a second lidar, a second second antenna array, a second multispectral camera, a second airborne processing unit, a GNSS module, and a 4G / 5G communication antenna. The second lidar performs real-time 3D scanning of the forest canopy and constructs its 3D structure in real time based on the scan results. The second multispectral camera acquires the spectral information of the forest canopy. The second airborne processing unit stores the 3D structure and spectral information of the forest canopy in real time and transmits them to the 4G / 5G communication antenna. The GNSS module acquires the spatial information of the forest-based UAV in real time and transmits it to the 4G / 5G communication antenna. The first antenna array receives real-time video streams of the forest understory environment, spectral images of forest understory vegetation, and 3D point clouds of the forest understory. It also forwards the received information to the 4G / 5G communication antenna. The 4G / 5G communication antenna transmits the spatial information of the drone in the forest, the real-time video streams of the forest understory environment, the spectral images of forest understory vegetation, the 3D point clouds of the forest understory, the 3D structure of the forest canopy, and the spectral information of the forest canopy to the ground control base station. It also receives control commands from the ground control base station and sends the control commands to the second antenna array to control the drone in the forest. The second antenna array also sends the control commands and the real-time spatial information of the drone in the forest to the first antenna array to control the drone in the forest.

4. The forest three-dimensional survey device based on dual UAV collaboration according to claim 3, characterized in that: The ground control base station includes a communication ground station and a graphics workstation. The communication ground station is used to receive spatial information of the UAV above the forest, real-time video stream of the forest environment, spectral images of forest vegetation and 3D point cloud of forest, 3D structure of forest canopy and spectral information of forest canopy, and transmit the control commands issued by the operator through the control software to the UAV above the forest, and the UAV above the forest forwards them to the UAV below the forest, so as to realize the coordinated control of the two UAV platforms. The graphics workstation is used to receive spatial information of UAVs in the forest, real-time video streams of the forest environment, spectral images of forest vegetation and 3D point clouds of the forest, 3D structure of the forest canopy and spectral information of the forest canopy, forwarded by the communication ground station. Based on the received information, it monitors the working status of UAVs in the forest and UAVs in the forest in real time, and performs data fusion and modeling on the received information to generate and visualize a real forest scene model.

5. A method for three-dimensional forest survey based on dual-UAV collaboration, utilizing the three-dimensional forest survey device based on dual-UAV collaboration as described in claim 4, characterized in that: Specifically, the steps include the following: S1: The operator plans the target path for the understory drone on the ground control base station and makes the overstory drone automatically follow the movement of the understory drone. The first flight position of the overstory drone is directly above the canopy corresponding to the first path point of the understory drone. S2: The operator sends control commands to the UAV above the forest via the ground control base station. The UAV above the forest forwards the control commands to the UAV below the aircraft, and the UAV below the aircraft flies autonomously according to the target path. S3: The understory drone collects real-time video streams of the understory environment, spectral images of understory vegetation, and three-dimensional point clouds of the understory at time t, and transmits the collected information at time t back to the ground control base station via the understory drone. S4: Based on the comprehensive channel quality of the drones above and below the forest at time t, adjust the relative positions of the drones above and below the forest in real time to ensure optimal channel quality for both drones above and below the forest. S5: At time t, the UAV in the forest collects its own spatial information, the three-dimensional structure of the forest canopy, and the spectral information of the forest canopy, and transmits the collected information at time t back to the ground control base station. The ground control station performs data fusion and modeling on the collected results of the UAV under the forest and the UAV in the forest according to the time sequence to generate a real forest scene model. S6: Replace time t with time t+1, and repeat steps S3-S5 until the drone below the aircraft completes its flight mission along the target path. The operator controls the drone below the aircraft to return to its home base via the ground control base station. At the same time, the drone below the aircraft guides the drone on the aircraft back to its respective take-off and landing point.

6. The forest three-dimensional survey method based on dual UAV collaboration according to claim 5, characterized in that: Step S4 specifically includes the following steps: S41: Let the horizontal position of the UAV in the forest at time t be... The horizontal position of the drone under the forest at time t is The horizontal distance between the drones above and below the forest is The overall channel quality at time t is Q(t); S42: Calculate the expected communication distance at time t based on the comprehensive channel quality at time t: =clip( ( ), , ); in, Let be the expected communication distance at time t. To preset the optimal channel quality, Sensitivity coefficient and These represent the lower limit of the safe distance and the upper limit of the effective communication distance, respectively. `clip(x,a,b)` means restricting the variable `x` to the interval [a,b], where `a` and `b` are the upper and lower limits of the interval, respectively. To preset the theoretically optimal communication distance; S43: Construct the joint cost function and solve for the optimal adjustment using gradient descent: ； ； ； ； in, To achieve the optimal adjustment amount, For the joint cost function, For distance error term, This is the channel quality prediction error term. For dynamic weighting coefficients, Adjust the vector for position. The regularization coefficient is . ( This is a local channel quality spatial prediction model based on historical and real-time detection data. To preset the optimal channel quality; S44: Based on the optimal adjustment amount, the target position of the drone above the forest is calculated using the following formula, and the drone above the forest is driven to fly to the target position based on the calculation result, thus completing the adjustment of the relative position between the drone above the forest and the drone below the forest: ； in, The target location for the drone in the forest.

Images