Forestry accurate measurement system based on unmanned aerial vehicle laser radar

Abstract

The invention relates to the technical field of forestry management, in particular to a forestry accurate measurement system based on an unmanned aerial vehicle laser radar, which comprises an unmanned aerial vehicle flight platform, a laser radar sensor, a data acquisition module, a data transmission module, a data processing module and a user interaction module, the unmanned aerial vehicle flying platform is responsible for carrying a laser radar sensor to shuttle over a forest region, and the flying capability of the unmanned aerial vehicle flying platform is utilized; the measurement precision is high, the system based on the unmanned aerial vehicle laser radar can penetrate through crown gap canopies, capture fine structures of tree branches and under-forest terrains and obtain three-dimensional space data with millimeter-level precision by utilizing the high-precision distance measurement characteristics of laser beams and point clouds, and the three-dimensional space data can be used for measuring the height of a single tree, the diameter at breast height and other parameters of the single tree, such as the height of the tree and the diameter at breast height. Compared with a traditional means, the method has the advantages that the error rate is greatly reduced, the accuracy is improved, the investment of manpower and material resources is reduced, the efficiency is improved, and a data foundation is built for accurate evaluation of forestry resources.

Description

Technical Field

[0001] The present invention relates to the technical field of forestry management, and in particular to a forestry precision measurement system based on unmanned aerial vehicle laser radar. Background Art

[0002] Forestry resources are a key component of the earth's ecosystem and play an indispensable role in regulating climate, conserving water and soil, and maintaining biodiversity. Traditional forestry measurement methods face many difficulties: manual field measurement is not only labor-intensive and inefficient, but also limited by complex terrain and dense vegetation, making it difficult for personnel to reach some areas, resulting in incomplete data collection. For example, in deep mountains and dense forests, investigators need to travel on foot, which consumes a lot of time and physical strength, but may still miss some tree information in hidden areas.

[0003] Ground optical measuring instruments, such as total stations and levels, have reasonable accuracy, but their field of view is limited, making it difficult to obtain overall information about large areas of forests and to quickly outline the macroscopic features of forests. Aerial photogrammetry relies on manned aircraft, which is expensive, has limited flight plans, and low shooting frequency. It is difficult to meet the needs of dynamic monitoring of forestry resources, and is also slightly insufficient in the detailed depiction of the vertical structure of trees and the topography under the forest.

[0004] Although satellite remote sensing technology can cover large areas, it is difficult to accurately identify the growth parameters of individual trees due to resolution issues. The information provided is relatively general and has limitations in forestry refined management and precise resource assessment. For this reason, a forestry precision measurement system based on UAV lidar is proposed. Summary of the invention

[0005] In view of this, the present invention provides a forestry precision measurement system based on UAV laser radar to solve or alleviate the technical problems existing in the prior art and at least provide a beneficial option.

[0006] The technical solution of the present invention is implemented as follows: a forestry precision measurement system based on a UAV laser radar comprises a UAV flight platform, a laser radar sensor, a data acquisition module, a data transmission module, a data processing module and a user interaction module;

[0007] The UAV flight platform is responsible for carrying the laser radar sensor to shuttle over the forest area. With its flight capability, the UAV flight route is set according to different terrains, different tree species, different planting years, and different growth conditions. A large amount of data is collected at different luminosity, different heights, different speeds, different widths, different wind directions, and wind speeds. The airborne laser radar covers the target forest area, creating spatial movement conditions for data collection. The point cloud data and image data of the laser radar are imported into the system for modeling and analysis to obtain results, ensuring that the measurement operation can reach forestry resources in different locations.

[0008] The laser radar sensor actively emits laser beams and receives reflected light signals. Based on the principle of light flight time, it accurately obtains three-dimensional spatial information of the forest area and generates massive point cloud data.

[0009] The data acquisition module is responsible for synchronously collecting various raw data generated by laser radar detection and key parameters of the drone during flight, organizing these fragmented information and preparing basic materials for subsequent transmission and processing;

[0010] The data transmission module transmits the large amount of data collected from the air in the forest area to the ground data processing center safely, quickly and stably through wireless communication networks and other means, thus opening up the link between aerial collection and ground processing;

[0011] The data processing module receives the transmitted raw data, performs pre-processing and optimization, and then performs in-depth mining and analysis to extract accurate measurement results such as tree height and diameter at breast height, and can also analyze the dynamic changes of forest resources and convert the raw data into usable forestry information;

[0012] The user interaction module, on the one hand, allows operators to easily control drones and lidars to carry out measurement tasks; on the other hand, it intuitively displays the processed measurement results to forestry personnel.

[0013] Further preferably, the UAV flight platform selects a suitable body according to the scope of the measured forest area and the complexity of the terrain. Small multi-rotors are used for small mountain forest areas, and large fixed-wings are used for vast plain forests. The flight control system integrates a high-precision inertial measurement unit and a global positioning system. The flight controller receives instructions, adjusts the attitude, speed and route of the UAV, and the lidar scans according to a preset trajectory; the UAV-mounted lidar obtains point cloud data, including determining single tree segmentation data, field personnel proofreading, determining the number of sample plots, advanced peeling through manual felling, advanced manual measurement, recording the volume of each tree and the size of each piece of wood, inputting into the system, and performing big data analysis, modeling proofreading and output A spreadsheet is used; first, the drone is equipped with a laser radar device to scan the forest area and obtain a large amount of point cloud data, which forms the basis for all subsequent work. After obtaining the point cloud data, the first key process is to carry out single tree segmentation data processing, and use professional algorithms to accurately distinguish the mixed point clouds into each independent tree, preliminarily outline the contour information of the individual trees, and lay the foundation for subsequent precise measurement. The proofreading link of field personnel follows closely. Experienced forestry staff go deep into the forest area to verify the segmented single tree data on the spot, correct the possible deviations of the algorithm, and ensure the accuracy of the data source. After all, the actual environment is complex and changeable, and some special landforms and tree growth trends are simply based on It is difficult to accurately grasp the situation by relying on algorithms, and it is also extremely important to determine the number of sample plots. According to the area of ​​the forest area, the complexity of the terrain, and the density of tree distribution, the number and range of representative sample plots are scientifically planned, and the situation of the entire forest area is estimated by the sample plot data. This can not only ensure the reliability of the data, but also improve the overall operation efficiency. After that, the advanced debarking process is entered. Through manual felling, field workers gradually remove the outer bark of the tree. The reason for using this seemingly traditional manual felling method is that the bark is irregular and the combination with the wood part is complex. Manual operation can minimize the interference with the measurement of the main body of the tree and obtain more accurate data on the wood part. In the advanced manual measurement stage The staff used professional measuring tools to record in detail the volume of each tree and the precise size of each piece of wood after it was cut down. These first-hand data are the key basis for subsequent analysis. After entering the system, big data analysis technology begins to play a role. Massive measurement data are imported into the system. After complex data mining and analysis algorithms, deep information such as tree growth patterns and timber stock distribution trends are excavated. At the same time, modeling and proofreading work is carried out simultaneously, and the acquired data is used to build a three-dimensional tree model, which is verified with the previous point cloud data and field measurements to check for omissions and make up for deficiencies, so that the model is infinitely close to the real tree state. Finally, the system will integrate and analyze the results and issue them as a well-organized spreadsheet.

[0014] Further preferably, the laser radar sensor selects a suitable laser radar according to the usage scenario. The pulse laser radar emits high-energy pulse laser, has a long optical range, is suitable for rapid large-area terrain mapping, and can penetrate a certain vegetation cover to obtain the terrain under the forest; the phase laser radar emits continuous laser, has extremely high accuracy, and is good at capturing the detailed structure of a single tree;

[0015] Point cloud data volume

[0016] 240,000-point scale: This level of point cloud data is suitable for small-scale, refined forestry operation scenarios. The relatively small amount of data means that the subsequent processing pressure is relatively small. It can quickly focus on the precise features of a single or a few trees, and accurately capture the texture of the trunk and the details of the branch forks.

[0017] 2 million point scale: A massive amount of 2 million point cloud data is used for large-scale forest resource surveys and overall planning of mountain forests. It can completely cover complex terrain and dense forest canopies, without missing any landform fluctuations or vegetation distribution differences, and build a very detailed three-dimensional forest model, providing comprehensive and three-dimensional data support for macro forestry decision-making;

[0018] Drone flight altitude

[0019] 240,000 points collected at corresponding altitude: At this time, the drone's flight altitude is relatively low, maintained in the range of 50-80 meters. Low-altitude flight can concentrate the energy of the laser beam emitted by the lidar, and the spot size is smaller. After hitting the surface of the tree, the reflected signal is strong and the accuracy is high.

[0020] 2 million points of acquisition corresponding to altitude: In order to obtain large-area data, the flight altitude is increased to 100-200 meters. High-altitude flight expands the field of view of the lidar, and the single scanning range is significantly increased, which can quickly cover large areas of mountains and forests;

[0021] Drone flight speed

[0022] 240,000-point acquisition speed: The drone flies at a slow speed of 2-3 meters per second. The slow speed gives the lidar enough time to process each laser pulse signal, from emission, reflection from the tree, to reception and analysis, to collect a very dense point cloud;

[0023] 2 million point acquisition speed: In response to the demand for large amounts of data, the flight speed is accelerated to 8-15 meters per second. This is because the higher flight altitude reduces the dependence on the accuracy of single point cloud acquisition. At the same time, the performance of the lidar has been optimized and adapted. When performing large-scale forest survey tasks, high-speed flying drones can cover long distances in a short time and quickly accumulate a large number of point clouds;

[0024] Drone flight distance

[0025] 240,000-point acquisition line spacing: The flight line spacing is set very narrow, at 10-50 meters. The compact line spacing ensures close connection between scanning strips, without data gaps, and achieves seamless coverage of small areas;

[0026] 2 million points of collection line spacing: For large-area operations, the flight line spacing is widened to 50-100 meters. The higher flight altitude is matched with the wide line spacing to ensure comprehensive coverage while avoiding data redundancy caused by repeated scanning of the same area.

[0027] Further preferably, the data acquisition module includes a synchronous trigger mechanism and a raw data cache. The synchronous trigger mechanism is closely linked with the UAV flight control system and the laser radar, and accurately triggers the laser radar to emit laser pulses according to the preset route and scanning frequency, and synchronously records the flight attitude, position, and timestamp information, so that the point cloud data can be attached with precise time and space labels. The raw data cache is equipped with a high-speed and large-capacity cache to temporarily store the time difference and angle of the laser radar reflected light, as well as the UAV attitude angle, longitude and latitude data.

[0028] Further preferably, the data transmission module includes a wireless communication network and data encoding and encryption. The wireless communication network utilizes links such as maritime satellites, Beidou satellites, 4G / 5G networks, etc. to transmit the collected data to the ground station in real time. The data encoding and encryption encodes the collected data according to a specific algorithm.

[0029] Further preferably, the data processing module includes data preprocessing, feature extraction and measurement, and change monitoring. The data preprocessing and denoising process filters out abnormal point clouds caused by flying birds, clouds, and equipment noise interference; the coordinate conversion converts the local coordinates based on the drone into the geodetic coordinate system to unify the data benchmark; the point cloud splicing integrates multiple scans and multi-angle data to form a complete three-dimensional point cloud of the forest area; the feature extraction and measurement, based on the point cloud, uses machine learning and deep learning algorithms to identify individual trees and extract geometric features such as tree height, breast diameter, and crown width; constructs a digital terrain model (DTM) and a digital surface model (DSM), calculates terrain parameters such as forest slope and aspect, and estimates forest stock; the change monitoring compares the point cloud data collected at different times, analyzes dynamic conditions such as tree growth, forest area changes, and vegetation coverage changes, and provides a basis for forest resource management decisions.

[0030] Further preferably, the user interaction module includes a control terminal and a display platform. The control terminal equips the operator with a handheld remote control, a tablet computer, etc., has an intuitive interface, and monitors the flight attitude and equipment status in real time. The display platform is used to display the processed measurement results through a laptop computer or a large-screen display. The two-dimensional forestry map marks the location and attributes of trees, and the three-dimensional model intuitively presents the three-dimensional structure of the forest, and the measurement report can be consulted.

[0031] Further preferably, the following is the formula for estimating forest volume:

[0032]

[0033] Calculate the volume of a single tree V using LiDAR data i , and then add up the timber volume of all individual trees to get the forest volume V. There are many empirical models for calculating the timber volume of individual trees. For example, for coniferous trees, V is often used. i =a+bD 2 H, D are the diameter at breast height, H is the tree height, and a and b are parameters determined according to the tree species and region.

[0034] The embodiment of the present invention has the following advantages due to the adoption of the above technical solution:

[0035] 1. The measurement accuracy of the present invention has been greatly improved. The system based on UAV lidar utilizes the high-precision ranging characteristics of laser beams and point clouds, which can penetrate the canopy gaps between tree crowns, capture the subtle structure of tree branches and understory terrain, and obtain three-dimensional spatial data with millimeter-level accuracy. Whether it is single tree parameters such as tree height and breast diameter, or the slope and slope direction of complex terrain, the measurement error rate is greatly reduced compared with traditional methods, thereby improving accuracy, reducing the investment of manpower and material resources, and improving efficiency, laying a solid data foundation for the accurate assessment of forestry resources.

[0036] 2. Compared with the traditional manual tree-by-tree measurement and lengthy aerial photography flight preparation, the UAV in this invention has strong maneuverability, and its takeoff and landing are not restricted by terrain. It can quickly fly to the target forest area, cover a large area in a short time, and collect massive data at one time, which greatly compresses the measurement cycle. It is especially suitable for large-scale and urgent forestry measurement projects, such as rapid assessment of resource losses after sudden forest fires.

[0037] 3. With the flexible deployment of drones, the present invention can repeatedly measure the same forest area at different time periods as needed, accumulate multi-temporal data, and compare and analyze these data. Forestry workers can clearly understand the growth rate of trees, the increase and decrease of forest area, and the changes in vegetation coverage, providing a dynamic basis for long-term health monitoring of forest ecosystems and the formulation of sustainable development strategies.

[0038] 4. The point cloud data generated by the lidar in the present invention contains rich information. It can not only present the geometric appearance of the forest, but also combine with algorithms to analyze key indicators such as forest density, stand density, and stock volume, outlining the overall picture of the forest in multiple dimensions, overcoming the limitations of traditional measurement of single-dimensional data, and making forestry resource management decisions more scientific.

[0039] 5. During the entire measurement process of the present invention, the operator can remotely control the drone operation without going deep into the high-risk areas of the forest, avoiding the potential safety risks brought by encountering wild animals, bad weather, and complex terrain, ensuring the safety of the staff and reducing the operation cost.

[0040] The above summary is for illustrative purposes only and is not intended to be limiting in any way. In addition to the illustrative aspects, embodiments and features described above, further aspects, embodiments and features of the present invention will be readily apparent by reference to the accompanying drawings and the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0042] Figure 1 This is a flowchart for obtaining point cloud data of the present invention;

[0043] Figure 2 is a processing flow chart of point cloud data of the present invention;

[0044] Figure 3 A demonstration diagram of obtaining point cloud data for the UAV airborne laser radar of the present invention;

[0045] Figure 4 It is a demonstration diagram of plant number identification of the present invention;

[0046] Figure 5 It is a demonstration diagram of plant number coordinate identification of the present invention;

[0047] Figure 6 This is a demonstration diagram for plant number identification of the present invention. DETAILED DESCRIPTION

[0048] In the following, only some exemplary embodiments are briefly described. As those skilled in the art will appreciate, the described embodiments may be modified in various ways without departing from the spirit or scope of the present invention. Therefore, the drawings and descriptions are considered to be exemplary and non-restrictive in nature.

[0049] The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0050] like Figure 1-6As shown, an embodiment of the present invention provides a forestry precision measurement system based on a UAV laser radar, including a UAV flight platform, a laser radar sensor, a data acquisition module, a data transmission module, a data processing module and a user interaction module;

[0051] The UAV flight platform is responsible for carrying the LiDAR sensor to shuttle over the forest area. With its flying ability, the UAV flight route is set according to different terrains, different tree species, different planting years, and different growth conditions. A large amount of data is collected at different luminosity, different heights, different speeds, different widths, different wind directions, and wind speeds. The airborne LiDAR covers the target forest area, creating spatial movement conditions for data collection. The point cloud data and image data of the LiDAR are imported into the system for modeling and analysis to obtain results, ensuring that the measurement operation can reach forestry resources in different locations.

[0052] The laser radar sensor actively emits laser beams and receives reflected light signals. Based on the principle of light flight time, it accurately obtains the three-dimensional spatial information of the forest area and generates massive point cloud data.

[0053] The data acquisition module is responsible for synchronously collecting various raw data generated by lidar detection and key parameters of the drone during flight, organizing these fragmented information and preparing basic materials for subsequent transmission and processing;

[0054] The data transmission module transmits the large amount of data collected from the air in the forest area to the ground data processing center safely, quickly and stably through wireless communication networks and other means, thus opening up the link between aerial collection and ground processing;

[0055] The data processing module receives the transmitted raw data, performs pre-processing and optimization, and then conducts in-depth mining and analysis to extract accurate measurement results such as tree height and diameter at breast height. It can also analyze the dynamic changes of forest resources and convert the raw data into usable forestry information.

[0056] The user interaction module, on the one hand, allows operators to easily control drones and lidars to carry out measurement tasks; on the other hand, it intuitively displays the processed measurement results to forestry personnel.

[0057] In one embodiment, the UAV flight platform selects a suitable body according to the scope of the measured forest area and the complexity of the terrain. Small multi-rotors are used for small mountain forest areas, and large fixed-wings are used for vast plain forests. The flight control system integrates a high-precision inertial measurement unit and a global positioning system. The flight controller receives instructions, adjusts the attitude, speed and route of the UAV, and the laser radar scans according to a preset trajectory; the UAV-mounted laser radar obtains point cloud data, including determining the single tree segmentation data, field personnel proofreading, determining the number of sample plots, advanced peeling through manual felling, advanced manual measurement, recording the volume of each tree and the size of each piece of wood, inputting into the system, and performing big data analysis, modeling proofreading and output A spreadsheet is used; first, the drone is equipped with a laser radar device to scan the forest area and obtain a large amount of point cloud data, which forms the basis for all subsequent work. After obtaining the point cloud data, the first key process is to carry out single tree segmentation data processing, and use professional algorithms to accurately distinguish the mixed point clouds into each independent tree, preliminarily outline the contour information of the individual trees, and lay the foundation for subsequent precise measurement. The proofreading link of field personnel follows closely. Experienced forestry staff go deep into the forest area to verify the segmented single tree data on the spot, correct the possible deviations of the algorithm, and ensure the accuracy of the data source. After all, the actual environment is complex and changeable, and some special landforms and tree growth trends are simply based on It is difficult to accurately grasp the situation by relying on algorithms, and it is also extremely important to determine the number of sample plots. According to the area of ​​the forest area, the complexity of the terrain, and the density of tree distribution, the number and range of representative sample plots are scientifically planned, and the situation of the entire forest area is estimated by the sample plot data. This can not only ensure the reliability of the data, but also improve the overall operation efficiency. After that, the advanced debarking process is entered. Through manual felling, field workers gradually remove the outer bark of the tree. The reason for using this seemingly traditional manual felling method is that the bark is irregular and the combination with the wood part is complex. Manual operation can minimize the interference with the measurement of the main body of the tree and obtain more accurate data on the wood part. In the advanced manual measurement stage The staff used professional measuring tools to record in detail the volume of each tree and the precise size of each piece of wood after it was cut down. These first-hand data are the key basis for subsequent analysis. After entering the system, big data analysis technology begins to play a role. Massive measurement data are imported into the system. After complex data mining and analysis algorithms, deep information such as tree growth patterns and timber stock distribution trends are excavated. At the same time, modeling and proofreading work is carried out simultaneously, and the acquired data is used to build a three-dimensional tree model, which is verified with the previous point cloud data and field measurements to check for omissions and make up for deficiencies, so that the model is infinitely close to the real tree state. Finally, the system will integrate and analyze the results and issue them as a well-organized spreadsheet.

[0058] In one embodiment, a laser radar sensor is selected according to the usage scenario. The pulse laser radar emits high-energy pulse lasers with a long optical range, which is suitable for rapid large-area terrain mapping and can penetrate certain vegetation cover to obtain forest terrain. The phase laser radar emits continuous lasers with extremely high accuracy and is good at capturing the detailed structure of a single tree.

[0059] Point cloud data volume

[0060] 240,000-point scale: This level of point cloud data is suitable for small-scale, refined forestry operation scenarios. The relatively small amount of data means that the subsequent processing pressure is relatively small. It can quickly focus on the precise features of a single or a few trees, and accurately capture the texture of the trunk and the details of the branch forks.

[0061] 2 million point scale: A massive amount of 2 million point cloud data is used for large-scale forest resource surveys and overall planning of mountain forests. It can completely cover complex terrain and dense forest canopies, without missing any landform fluctuations or vegetation distribution differences, and build a very detailed three-dimensional forest model, providing comprehensive and three-dimensional data support for macro forestry decision-making;

[0062] Drone flight altitude

[0063] 240,000 points collected at corresponding altitude: At this time, the drone's flight altitude is relatively low, maintained in the range of 50-80 meters. Low-altitude flight can concentrate the energy of the laser beam emitted by the lidar, and the spot size is smaller. After hitting the surface of the tree, the reflected signal is strong and the accuracy is high.

[0064] 2 million points of acquisition corresponding to altitude: In order to obtain large-area data, the flight altitude is increased to 100-200 meters. High-altitude flight expands the field of view of the lidar, and the single scanning range is significantly increased, which can quickly cover large areas of mountains and forests;

[0065] Drone flight speed

[0066] 240,000-point acquisition speed: The drone flies at a slow speed of 2-3 meters per second. The slow speed gives the lidar enough time to process each laser pulse signal, from emission, reflection from the tree, to reception and analysis, to collect a very dense point cloud;

[0067] 2 million point acquisition speed: In response to the demand for large amounts of data, the flight speed is accelerated to 8-15 meters per second. This is because the higher flight altitude reduces the dependence on the accuracy of single point cloud acquisition. At the same time, the performance of the lidar has been optimized and adapted. When performing large-scale forest survey tasks, high-speed flying drones can cover long distances in a short time and quickly accumulate a large number of point clouds;

[0068] Drone flight distance

[0069] 240,000-point acquisition line spacing: The flight line spacing is set very narrow, at 10-50 meters. The compact line spacing ensures close connection between scanning strips, without data gaps, and achieves seamless coverage of small areas;

[0070] 2 million points of collection line spacing: For large-area operations, the flight line spacing is widened to 50-100 meters. The higher flight altitude is matched with the wide line spacing to ensure comprehensive coverage while avoiding data redundancy caused by repeated scanning of the same area.

[0071] In one embodiment, the data acquisition module includes a synchronous trigger mechanism and a raw data cache. The synchronous trigger mechanism is closely linked with the UAV flight control system and the laser radar. According to the preset route and scanning frequency, it accurately triggers the laser radar to emit laser pulses, and synchronously records the flight attitude, position, and timestamp information, so that the point cloud data can be attached with precise time and space labels. The raw data cache is equipped with a high-speed, large-capacity cache to temporarily store the time difference and angle of the laser radar reflected light, as well as the UAV attitude angle, longitude and latitude data.

[0072] In one embodiment, the data transmission module includes a wireless communication network and data encoding and encryption. The wireless communication network uses links such as maritime satellites, Beidou satellites, 4G / 5G networks, etc. to transmit the collected data to the ground station in real time. The data encoding and encryption encodes the collected data according to a specific algorithm.

[0073] In one embodiment, the data processing module includes data preprocessing, feature extraction and measurement, and change monitoring. Data preprocessing, denoising processing filters out abnormal point clouds caused by flying birds, clouds, and equipment noise interference; coordinate conversion converts local coordinates based on drones into geodetic coordinate systems to unify data benchmarks; point cloud splicing integrates multiple scans and multi-angle data to form a complete three-dimensional point cloud of the forest area, feature extraction and measurement, based on point clouds, use machine learning and deep learning algorithms to identify individual trees, and extract geometric features such as tree height, breast diameter, and crown width; construct digital terrain models (DTM) and digital surface models (DSM), calculate terrain parameters such as forest slope and slope aspect, estimate forest stock, change monitoring, compare point cloud data collected at different times, analyze dynamic conditions such as tree growth, forest area changes, and vegetation coverage changes, and provide a basis for forest resource management decisions.

[0074] In one embodiment, the user interaction module includes a control terminal and a display platform. The control terminal equips the operator with a handheld remote control, a tablet computer, etc., with an intuitive interface, and monitors the flight attitude and equipment status in real time. The display platform is used to display the processed measurement results through a laptop computer and a large-screen display. The two-dimensional forestry map marks the location and attributes of trees, and the three-dimensional model intuitively presents the three-dimensional structure of the forest, and the measurement report can be consulted.

[0075] In one embodiment, the following is the formula used for forest stock estimation:

[0076]

[0077] Calculate the volume of a single tree V using LiDAR data i , and then add up the timber volume of all individual trees to get the forest volume V. There are many empirical models for calculating the timber volume of individual trees. For example, for coniferous trees, V is often used. i =a+bD 2 H, D are the diameter at breast height, H is the tree height, and a and b are parameters determined according to the tree species and region.

[0078] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of various changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. The forestry precision measurement system based on UAV laser radar is characterized by: It includes UAV flight platform, LiDAR sensor, data acquisition module, data transmission module, data processing module and user interaction module; The UAV flight platform is responsible for carrying the laser radar sensor to shuttle over the forest area. With its flight capability, the UAV flight route is set according to different terrains, different tree species, different planting years, and different growth conditions. A large amount of data is collected at different luminosity, different heights, different speeds, different widths, different wind directions, and wind speeds. The airborne laser radar covers the target forest area, creating spatial movement conditions for data collection. The point cloud data and image data of the laser radar are imported into the system for modeling and analysis to obtain results, ensuring that the measurement operation can reach forestry resources in different locations. The laser radar sensor actively emits laser beams and receives reflected light signals. Based on the principle of light flight time, it accurately obtains three-dimensional spatial information of the forest area and generates massive point cloud data. The data acquisition module is responsible for synchronously collecting various raw data generated by laser radar detection and key parameters of the drone during flight, organizing these fragmented information and preparing basic materials for subsequent transmission and processing; The data transmission module transmits the large amount of data collected from the air in the forest area to the ground data processing center safely, quickly and stably through wireless communication networks and other means, thus opening up the link between aerial collection and ground processing; The data processing module receives the transmitted raw data, performs pre-processing and optimization, and then performs in-depth mining and analysis to extract accurate measurement results such as tree height and diameter at breast height, and can also analyze the dynamic changes of forest resources and convert the raw data into usable forestry information; The user interaction module, on the one hand, allows operators to easily control drones and lidars to carry out measurement tasks; on the other hand, it intuitively displays the processed measurement results to forestry personnel.

2. The forestry precision measurement system based on UAV laser radar according to claim 1 is characterized in that: The UAV flight platform selects a suitable body according to the scope of the measured forest area and the complexity of the terrain. Small multi-rotors are used for small mountain forest areas, and large fixed-wings are used for vast plain forests. The flight control system integrates a high-precision inertial measurement unit and a global positioning system. The flight controller receives instructions, adjusts the attitude, speed and route of the UAV, and the laser radar scans according to a preset trajectory; the UAV-mounted laser radar obtains point cloud data, including determining single tree segmentation data, field personnel proofreading, determining the number of sample plots, advanced peeling through manual felling, advanced manual measurement, recording the volume of each tree and the size of each piece of wood, inputting into the system, and performing big data analysis, modeling proofreading and issuing electronic spreadsheets. ; First, the drone is equipped with a laser radar device to scan the forest area and obtain a large amount of point cloud data, which forms the basis for all subsequent work. After obtaining the point cloud data, the first key process is to carry out single tree segmentation data processing, and use professional algorithms to accurately distinguish the mixed point clouds into each independent tree, preliminarily outline the contour information of the individual trees, and lay the foundation for subsequent precise measurement. The proofreading link of field personnel follows closely. Experienced forestry staff go deep into the forest area to verify the segmented single tree data on the spot, correct possible deviations in the algorithm, and ensure the accuracy of the data source. After all, the actual environment is complex and changeable, and some special landforms and tree growth trends rely solely on algorithms. It is difficult to grasp accurately, and determining the number of sample plots is also extremely critical. According to the area of ​​the forest area, the complexity of the terrain, and the density of tree distribution, the number and range of representative sample plots are scientifically planned, and the sample plot data are used to infer the situation of the entire forest area, which can not only ensure the reliability of the data, but also improve the overall operation efficiency. Then enter the advanced debarking process. Through manual felling, field workers gradually remove the outer bark of the tree. The reason for using this seemingly traditional manual felling method is that the bark is irregular and the combination with the wood part is complex. Manual operation can minimize the interference with the measurement of the main body of the tree and obtain more accurate data on the wood part. In the advanced manual measurement stage, The staff used professional measuring tools to record in detail the volume of each tree and the precise size of each piece of wood after it was cut down. These first-hand data are the key basis for subsequent analysis. After entering the system, big data analysis technology begins to play a role. Massive measurement data are imported into the system. After complex data mining and analysis algorithms, deep information such as tree growth patterns and timber stock distribution trends are excavated. At the same time, modeling and proofreading work is carried out simultaneously, and the acquired data is used to build a three-dimensional tree model, which is verified with the previous point cloud data and field measurements to check for omissions and make up for deficiencies, so that the model is infinitely close to the real tree state. Finally, the system will issue the results of the integration and analysis as a well-organized spreadsheet.

3. The forestry precision measurement system based on UAV laser radar according to claim 1 is characterized by: The laser radar sensor is selected according to the usage scenario. The pulse laser radar emits high-energy pulse laser, has a long optical range, is suitable for rapid large-area terrain mapping, and can penetrate a certain amount of vegetation to obtain the terrain under the forest; the phase laser radar emits continuous laser, has extremely high accuracy, and is good at capturing the detailed structure of a single tree; Point cloud data volume 240,000-point scale: This level of point cloud data is suitable for small-scale, refined forestry operation scenarios. The relatively small amount of data means that the subsequent processing pressure is relatively small. It can quickly focus on the precise features of a single or a few trees, and accurately capture the texture of the trunk and the details of the branch forks. 2 million point scale: A massive amount of 2 million point cloud data is used for large-scale forest resource surveys and overall planning of mountain forests. It can completely cover complex terrain and dense forest canopies, without missing any landform fluctuations or vegetation distribution differences, and build a very detailed three-dimensional forest model, providing comprehensive and three-dimensional data support for macro forestry decision-making; Drone flight altitude 240,000 points collected at corresponding altitude: At this time, the drone's flight altitude is relatively low, maintained in the range of 50-80 meters. Low-altitude flight can concentrate the energy of the laser beam emitted by the lidar, and the spot size is smaller. After hitting the surface of the tree, the reflected signal is strong and the accuracy is high. 2 million points of acquisition corresponding to altitude: In order to obtain large-area data, the flight altitude is increased to 100-200 meters. High-altitude flight expands the field of view of the lidar, and the single scanning range is significantly increased, which can quickly cover large areas of mountains and forests; Drone flight speed 240,000-point acquisition speed: The drone flies at a slow speed of 2-3 meters per second. The slow speed gives the lidar enough time to process each laser pulse signal, from emission, reflection from the tree, to reception and analysis, to collect a very dense point cloud; 2 million point acquisition speed: In response to the demand for large amounts of data, the flight speed is accelerated to 8-15 meters per second. This is because the higher flight altitude reduces the dependence on the accuracy of single point cloud acquisition. At the same time, the performance of the lidar has been optimized and adapted. When performing large-scale forest survey tasks, high-speed flying drones can cover long distances in a short time and quickly accumulate a large number of point clouds; Drone flight distance 240,000-point acquisition line spacing: The flight line spacing is set very narrow, at 10-50 meters. The compact line spacing ensures close connection between scanning strips, without data gaps, and achieves seamless coverage of small areas; 2 million points of collection line spacing: For large-area operations, the flight line spacing is widened to 50-100 meters. The higher flight altitude is matched with the wide line spacing to ensure comprehensive coverage while avoiding data redundancy caused by repeated scanning of the same area.

4. The forestry precision measurement system based on UAV laser radar according to claim 1 is characterized in that: The data acquisition module includes a synchronous trigger mechanism and a raw data cache. The synchronous trigger mechanism is closely linked with the UAV flight control system and the laser radar. According to the preset route and scanning frequency, it accurately triggers the laser radar to emit laser pulses, and synchronously records the flight attitude, position, and timestamp information, so that the point cloud data can be attached with precise time and space labels. The raw data cache is equipped with a high-speed and large-capacity cache to temporarily store the time difference and angle of the laser radar reflected light, as well as the UAV attitude angle, longitude and latitude and other data.

5. The forestry precision measurement system based on UAV laser radar according to claim 1 is characterized in that: The data transmission module includes a wireless communication network and data encoding and encryption. The wireless communication network uses links such as maritime satellites, Beidou satellites, 4G / 5G networks, etc. to transmit the collected data to the ground station in real time. The data encoding and encryption encodes the collected data according to a specific algorithm.

6. The forestry precision measurement system based on UAV laser radar according to claim 1 is characterized by: The data processing module includes data preprocessing, feature extraction and measurement, and change monitoring. The data preprocessing and denoising process filters out abnormal point clouds caused by flying birds, clouds, fog, and equipment noise interference; Coordinate transformation converts the local coordinates based on the drone into the geodetic coordinate system to unify the data benchmark; Point cloud stitching integrates multiple scans and multi-angle data to form a complete three-dimensional point cloud of the forest area. The feature extraction and measurement uses machine learning and deep learning algorithms based on point clouds to identify individual trees and extract geometric features such as tree height, breast diameter, and crown width; Construct digital terrain models (DTM) and digital surface models (DSM), calculate terrain parameters such as forest slope and aspect, estimate forest stock, monitor changes, compare point cloud data collected at different times, analyze dynamic conditions such as tree growth, forest area changes, and vegetation coverage changes, and provide a basis for forest resource management decisions.

7. The forestry precision measurement system based on UAV laser radar according to claim 1 is characterized by: The user interaction module includes a control terminal and a display platform. The control terminal is equipped with a handheld remote control, a tablet computer, etc. for the operator. The interface is intuitive and the flight attitude and equipment status are monitored in real time. The display platform is used to display the processed measurement results through a laptop computer and a large-screen display. The two-dimensional forestry map marks the location and attributes of trees, and the three-dimensional model intuitively presents the three-dimensional structure of the forest, and the measurement report is consulted.

8. The forestry precision measurement system based on UAV laser radar according to claim 1 is characterized by: The following is the formula used for estimating forest stock volume: Calculate the volume of a single tree V using LiDAR data i , and then add up the timber volume of all individual trees to get the forest volume V. There are many empirical models for calculating the timber volume of individual trees. For example, for coniferous trees, V is often used. i ＝a+ B 2 H, D are the diameter at breast height, H is the tree height, and a and b are parameters determined according to the tree species and region.

Images