Forest resource monitoring system based on airborne laser radar data

Abstract

The invention relates to the technical field of forest monitoring, in particular to a forest resource monitoring system based on airborne laser radar data, which is provided with a forest feature acquisition module, an under-forest feature analysis module, a feature preprocessing module and an under-forest monitoring adjustment module. The method comprises the following steps: screening feature monitoring sub-regions through an under-forest feature analysis module, marking tree points through a feature preprocessing module to determine a growth coefficient of a tree, or judging whether feature overlapping exists in a marked tree or not, obtaining crown morphological parameters, selecting an adjustment mode of an airborne laser radar through an under-forest monitoring adjustment module, and further obtaining a tree crown morphological parameter. The method achieves the quick recognition of the region affected by the laser pulse in the monitoring process, adaptively adjusts the monitoring mode of the laser radar according to the morphological characteristics of the upper vegetation, and improves the monitoring precision of forest resources.

Description

Technical Field

[0001] The present invention relates to the technical field of forest monitoring, and particularly to a forest resource monitoring system based on airborne lidar data. Background Art

[0002] Forest resources play an important role in maintaining ecological balance, providing ecological services, promoting economic development, etc. It is crucial to accurately and timely grasp the current situation and changes of forest resources. Traditional forest resource monitoring often relies on ground surveys, which are time-consuming and laborious, and difficult to implement in areas with complex terrain and inconvenient transportation. Although the monitoring method based on remote sensing images can cover a large area, it is limited by resolution and spectral information and is difficult to accurately obtain the three-dimensional structure information of forests. Airborne lidar mounts the lidar system on an aircraft, can quickly obtain the three-dimensional spatial information of large areas of forests, is not restricted by terrain and vegetation coverage, and has high spatial resolution and vertical accuracy.

[0003] However, a forest is a complex ecosystem with a multi-layered vegetation structure. Upper-layer plants such as tall trees usually have dense tree canopies, and their branches and leaves are distributed densely. When airborne lidar emits laser pulses from the air to scan the forest, these upper-layer plants will affect the laser pulses, making only part of the laser energy reach the understory vegetation. The canopy shapes, branch and leaf densities, and canopy densities of different tree species are different, and the degrees of influence on the laser also vary. Therefore, improving the monitoring reliability of understory vegetation is a technical problem to be solved urgently.

[0004] For example, Chinese Patent Application Publication No.: CN119355750A, this invention discloses a lidar data processing method for complex forest environments. The original point cloud data of the target forest area is obtained through an airborne lidar system or a backpack lidar system, and preprocessing is performed to obtain preprocessed point cloud data. By filtering the ground points in the preprocessed point cloud data, non-ground point cloud data is obtained, and the ground objects in the non-ground point cloud data are classified to obtain ground object classification point cloud data; based on the ground object classification point cloud data, a three-dimensional canopy point cloud model is constructed, and based on the single-tree parameter extraction method, the single-tree spatial positions and key forestry parameters of tree individuals in the target forest area are obtained.

[0005] The following problems also exist in the prior art: The prior art does not consider the influence of upper-layer vegetation on laser pulses during airborne lidar scanning due to the multi-layered vegetation structure of the forest, and it is difficult to quickly identify the areas where laser pulses are affected during the monitoring process. It cannot adaptively adjust the monitoring method of the lidar according to the morphological characteristics of the upper-layer vegetation, affecting the monitoring accuracy of forest resources. Summary of the Invention

[0006] To this end, the present invention provides a forest resource monitoring system based on airborne lidar data to overcome the problems in the prior art that in the monitoring process, it is impossible to quickly identify the areas affected by laser pulses, and it is impossible to adaptively adjust the monitoring method of lidar according to the morphological characteristics of the upper-layer vegetation, which affects the monitoring accuracy of forest resources.

[0007] To achieve the above object, the present invention provides a forest resource monitoring system based on airborne lidar data, including: A forest feature acquisition module, which is used to control the airborne lidar to acquire the point cloud data of the forest area to be measured; An understory feature analysis module, which is connected to the forest feature acquisition module and is used to divide the forest area to be measured into several monitoring sub-areas, and determine the terrain parameters based on the point cloud data of each monitoring sub-area to screen the characteristic monitoring sub-areas; A feature preprocessing module, which is respectively connected to the forest feature acquisition module and the understory feature analysis module, and is used to mark several tree points based on the point cloud data of the characteristic monitoring sub-areas, determine the distribution coefficient according to the distribution of the tree points to select the growth coefficient of the characteristic monitoring sub-areas determined according to the point cloud data, or determine whether there is feature overlap of the marked trees according to the comparison between the position of the marked trees and the understory identification difference area, and obtain the crown morphological parameters of the marked trees with feature overlap; An understory monitoring adjustment module, which is respectively connected to the forest feature acquisition module and the feature preprocessing module, and is used to select the adjustment method of the airborne lidar as determining the scanning frequency of the airborne lidar according to the growth coefficient and the distribution coefficient, or determining the laser scanning wavelength for the airborne lidar to scan the marked trees with feature overlap according to the crown morphological parameters; Wherein, the understory identification difference area is determined according to the terrain parameters.

[0008] Furthermore, the understory feature analysis module is used to determine the terrain parameters, wherein, The understory feature analysis module is used to construct a real-time terrain model of the monitoring sub-area according to the point cloud data; To divide the real-time terrain model into several sub-model areas, divide the terrain model samples into several sample sub-model areas, and establish an association relationship between the sub-model areas and the sample sub-model areas; To calculate the morphological coincidence degree between the sub-model area and the sample sub-model area respectively, and determine the minimum value of the morphological coincidence degree as the terrain parameter of the monitoring sub-area; The terrain model samples are determined according to the historical monitoring data of the monitoring sub-areas.

[0009] Furthermore, the understory feature analysis module is used to screen the characteristic monitoring sub-areas, wherein, If the terrain parameters of the monitored sub-region meet the characteristic determination conditions, the understory characteristic analysis module will screen the monitored sub-region as a characteristic monitoring sub-region; The characteristic determination condition is that the terrain parameter does not exceed a preset terrain threshold.

[0010] Further, the feature preprocessing module is used to mark a number of tree points and determine the distribution coefficient. Among them, The feature preprocessing module determines the height values of a number of preset collection points in the geodetic coordinate system within the feature monitoring sub-region according to the point cloud data, and marks the collection points whose height values exceed the preset height threshold as the tree points; The feature preprocessing module calculates the minimum interval distance between any one feature tree point and the remaining tree points within the feature monitoring sub-region, and determines the minimum interval distance as the distribution factor of the feature tree point; Calculate the variance of the distribution factors of a number of feature tree points, and determine the variance of the distribution factors as the distribution coefficient of the feature monitoring sub-region.

[0011] Further, if the distribution coefficient of the feature monitoring sub-region meets the uniform distribution determination conditions, the feature preprocessing module selects and determines the growth coefficients of a number of trees within the feature monitoring sub-region according to the point cloud data; If the distribution coefficient of the feature monitoring sub-region does not meet the uniform distribution determination conditions, the feature preprocessing module selects and determines whether there is feature overlap of the landmark trees according to the comparison between the location of the landmark trees and the understory recognition difference region, and obtains the crown shape parameters of the landmark trees with feature overlap; The uniform distribution determination condition is that the distribution coefficient does not exceed a preset distribution threshold.

[0012] Further, the feature preprocessing module is used to determine the growth coefficients of a number of trees within the feature monitoring sub-region. Among them, The feature preprocessing module determines the diameters of the trees at different height levels according to the point cloud data, determines the average diameter as the growth factor of the trees, calculates the variance according to the growth factors of a number of trees within the feature monitoring sub-region, and determines the variance as the growth coefficient.

[0013] Further, the feature preprocessing module is used to determine whether there is feature overlap of the landmark trees. Among them, The feature preprocessing module marks the feature tree points whose distribution factors of the feature tree points do not exceed the preset distribution reference value as landmark trees, and determines the area corresponding to the terrain parameters within the feature monitoring sub-region as the understory recognition difference region; If the marker trees are within the forest understory identification difference area, the feature preprocessing module determines that there is feature overlap between the marker trees and obtains tree crown morphology parameters.

[0014] Furthermore, it is characterized in that the feature preprocessing module is used to determine the crown morphological parameters, wherein: The feature preprocessing module obtains the crown point cloud of each marked tree, divides the crown point cloud into a number of voxels, determines the number of point clouds in the voxels, calculates the variance according to the number of point clouds in the voxels, and determines the variance as the crown morphological parameter.

[0015] Furthermore, the understory monitoring adjustment module is used to select an adjustment method for the airborne laser radar, wherein: If the growth coefficients of several trees in the characteristic monitoring sub-area are obtained, the understory monitoring adjustment module selects the adjustment method of the airborne laser radar to determine the scanning frequency of the airborne laser radar according to the growth coefficient and the distribution coefficient; If it is determined whether there is feature overlap among the marker trees and the crown morphological parameters are determined, the understory monitoring adjustment module selects an adjustment method for the airborne laser radar to determine the laser scanning wavelength for scanning each marker tree according to the crown morphological parameters.

[0016] Furthermore, the scanning frequency is negatively correlated with the growth coefficient and the distribution coefficient, respectively, and the laser scanning wavelength is positively correlated with the crown morphological parameters.

[0017] Compared with the prior art, the beneficial effect of the present invention lies in that the present invention sets a forest feature acquisition module, an understory feature analysis module, a feature preprocessing module, and an understory monitoring and adjustment module, obtains point cloud data of the forest area to be measured through the forest feature acquisition module, divides a number of monitoring sub-areas through the understory feature analysis module, and determines terrain parameters based on the point cloud data of each monitoring sub-area to screen the feature monitoring sub-area, marks a number of tree points through the feature preprocessing module, and selects the growth coefficients of a number of trees in the feature monitoring sub-area determined according to the point cloud data based on the distribution of the number of tree points, or determines whether there is feature overlap of the marked trees based on the comparison between the location of the marked trees and the understory identification difference area, and obtains the crown morphological parameters of the area with feature overlap, and selects the adjustment method of the airborne laser radar through the understory monitoring and adjustment module, thereby realizing the rapid identification of the area where the laser pulse is affected during the monitoring process, and adaptively adjusting the monitoring method of the laser radar according to the morphological characteristics of the upper vegetation, thereby improving the monitoring accuracy of forest resources.

[0018] In particular, the present invention determines topographic parameters based on the point cloud data of each monitoring sub-region through the underforest feature analysis module to screen out the feature monitoring sub-regions. It can be understood that the laser pulses of airborne lidar are easily affected by the upper-layer vegetation, thus affecting the accuracy of the point cloud data. A real-time topographic model is constructed based on the actually acquired point cloud data of the lidar. The accuracy of the currently acquired point cloud data of the lidar is characterized by the degree of coincidence between the real-time topographic model and the topographic model sample. The higher the degree of coincidence, the higher the accuracy of the point cloud data, and the smaller the influence of the upper-layer vegetation on the airborne lidar detection in the current monitoring sub-region. The lower the degree of coincidence, the lower the accuracy of the point cloud data, and the greater the influence of the upper-layer vegetation on the airborne lidar detection in the current monitoring sub-region. Screening out the regions with a large influence of the upper-layer vegetation on the airborne lidar detection, that is, the feature monitoring sub-regions, helps to take more targeted data collection for these regions in subsequent monitoring. Furthermore, it realizes the rapid identification of the regions where the laser pulses have an impact during the monitoring process, improving the monitoring accuracy of forest resources.

[0019] In particular, the present invention determines the distribution coefficient according to the distribution of several tree points through the feature preprocessing module to select the growth coefficient for determining the feature monitoring sub-region based on the point cloud data, or determines whether there is feature overlap of the landmark tree according to the comparison between the location of the landmark tree and the underforest recognition difference region, and obtains the crown shape parameters of the landmark tree with feature overlap. It can be understood that different tree distributions represent different forest structures and ecological environments. By choosing the appropriate method, the diversity of the forest can be monitored more carefully. Whether it is the overall characteristics of the evenly distributed region or the special characteristics of the locally concentrated region, accurate data collection can be obtained, and then the actual situation of the forest can be reflected more comprehensively and truly. Selecting different parameters according to different tree distributions helps to adopt more targeted monitoring measures for different regions. For the regions with evenly distributed trees, a relatively unified monitoring strategy can be carried out, while for the regions with uneven trees and local concentration, more targeted monitoring methods can be adopted, providing a more accurate basis for the scientific monitoring of forest resources. Furthermore, it realizes the rapid identification of the regions where the laser pulses have an impact during the monitoring process, adaptively adjusts the monitoring method of the lidar according to the morphological characteristics of the upper-layer vegetation, and improves the monitoring accuracy of forest resources.

[0020] In particular, under the condition that the distribution of tree points is relatively uniform, the growth coefficients of several trees in the feature monitoring sub-region are determined according to the point cloud data. It can be understood that when the tree distribution is uniform, the tree diameters obtained from the point cloud data can more accurately represent the size of the trees in the entire region. Based on these accurate diameter data, the degree of canopy shading of the understory vegetation can be more precisely evaluated. Because the uniform distribution of tree points makes the influence range of each tree relatively stable and predictable, unlike the uneven distribution where there are local dense shading or sparse non-shading situations, it can provide a more accurate basis for adjusting the lidar parameters to ensure that the laser can effectively penetrate the canopy gaps to reach the understory vegetation. Obtaining tree diameters and adjusting the lidar parameters under the condition of uniform tree point distribution helps to ensure the high reliability of the understory vegetation data obtained in the entire monitoring region. Furthermore, it realizes the rapid identification of the regions affected by laser pulses during the monitoring process, adaptively adjusts the monitoring method of the lidar according to the morphological characteristics of the upper-layer vegetation, and improves the monitoring accuracy of forest resources.

[0021] In particular, under the condition that the distribution of tree points is relatively uneven, it is determined whether there is feature overlap of the landmark trees according to the comparison between the position of the landmark trees and the understory recognition difference region, and the canopy morphological parameters of the region with feature overlap are obtained. It can be understood that the landmark trees are the trees with relatively concentrated tree point distribution, and the understory recognition difference region is the region where the terrain obtained by the lidar has a low coincidence degree with the historical terrain. The landmark trees are located in the understory recognition difference region, indicating the deviation of the terrain obtained by the airborne lidar in real time and the possibility of being affected by the upper-layer vegetation. Obtaining the canopy morphological parameters in these regions can evaluate the degree of canopy shading of the upper-layer canopy on the understory vegetation. The regular canopy shading is relatively uniform. Adjust the lidar parameters according to the actual shading situation to make the laser energy more reasonably distributed and improve the ability to penetrate the canopy to reach the understory vegetation. Furthermore, more reliable understory vegetation data can be obtained. Furthermore, it realizes the rapid identification of the regions affected by laser pulses during the monitoring process, adaptively adjusts the monitoring method of the lidar according to the morphological characteristics of the upper-layer vegetation, and improves the monitoring accuracy of forest resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 It is the structural block diagram of the forest resource monitoring system based on airborne lidar data in the embodiment of the present invention; Figure 2 It is the logical flow chart of the understory feature analysis module for screening the feature monitoring sub-region in the embodiment of the present invention; Figure 3 It is the logical flow chart of the feature preprocessing module for determining whether there is feature overlap of the landmark trees in the embodiment of the present invention; Figure 4This is a logic flowchart for selecting the adjustment method of the airborne lidar for the underforest monitoring and adjustment module in the embodiments of the present invention. Detailed implementation manners

[0023] In order to make the objectives and advantages of the present invention more clear and understandable, the present invention will be further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0024] The preferred implementation manners of the present invention will be described below with reference to the accompanying drawings. Those skilled in the art should understand that these implementation manners are only used to explain the technical principles of the present invention and do not limit the protection scope of the present invention.

[0025] It should be noted that in the description of the present invention, the terms indicating the direction or positional relationship such as "upper", "lower", "inner", "outer", etc. are based on the direction or positional relationship shown in the drawings. This is only for convenience of description and does not indicate or imply that the device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the present invention.

[0026] In addition, it should also be noted that in the description of the present invention, unless otherwise clearly specified and limited, the terms "installation" and "connection" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be directly connected or indirectly connected through an intermediate medium, and it can be the communication inside two elements. For those skilled in the art, the specific meanings of the above terms in the present invention can be understood according to specific situations.

[0027] Please refer to Figure 1 As shown, it is a structural block diagram of the forest resource monitoring system based on airborne lidar data in the embodiments of the present invention. The forest resource monitoring system based on airborne lidar data of the present invention includes: A forest feature acquisition module, which is used to control the airborne lidar to acquire the point cloud data of the forest area to be measured; Specifically, the present invention does not limit the specific structure of the forest feature acquisition module. Preferably, it can be a processor used by a computer to control the airborne lidar to acquire the point cloud data of the forest area to be measured, which will not be elaborated here.

[0028] An underforest feature analysis module, which is connected to the forest feature acquisition module and is used to divide the forest area to be measured into several monitoring sub-areas, and determine the terrain parameters based on the point cloud data of each monitoring sub-area to screen out the feature monitoring sub-areas; Specifically, the division size of the monitoring sub-region can be set by those skilled in the art according to the accuracy requirements of forest resource monitoring. The higher the accuracy requirements, the smaller the division size of the monitoring sub-region. Preferably, the division size of the monitoring sub-region can be 200m×200m.

[0029] Specifically, the airborne lidar is usually installed at specific parts of the aircraft. For example, it can be installed on the belly of the aircraft, under the wings, or at the tail of the aircraft. Preferably, it can be installed on the belly of the aircraft so that the laser beam is emitted vertically downward or nearly vertically downward, enabling direct acquisition of the vertical point cloud data of the target area, which is beneficial for the accurate measurement of information such as terrain and vegetation height. Moreover, the space on the belly of the aircraft is relatively large, facilitating the installation and fixation of the lidar equipment, and at the same time, it can better protect the equipment from airflow interference. Details are not elaborated here.

[0030] Specifically, the lidar system emits laser pulses to the forest area to be measured, digitally processes the received reflected light information, and calculates the coordinates of each reflection point in three-dimensional space based on the flight time of the laser pulse and the position and attitude information of the aircraft, thereby generating the point cloud data of the forest area. Details are not elaborated here.

[0031] Specifically, the present invention does not limit the specific structure of the understory feature analysis module. Preferably, it can be a microprocessor for dividing the monitoring sub-region, determining the terrain parameters, and screening the feature monitoring sub-region. Details are not elaborated here.

[0032] The feature preprocessing module is respectively connected to the forest feature acquisition module and the understory feature analysis module, and is used to mark a number of tree points based on the point cloud data of the feature monitoring sub-region, determine the distribution coefficient according to the distribution of the tree points to select the growth coefficient of the feature monitoring sub-region determined according to the point cloud data, or determine whether there is feature overlap of the marked trees according to the comparison between the position of the marked trees and the understory recognition difference region, and obtain the crown shape parameters of the marked trees with feature overlap; Specifically, the present invention does not limit the specific structure of the feature preprocessing module. Preferably, it can be a programmable logic controller for marking a number of tree points, determining the distribution coefficient, determining the growth coefficient, and determining whether there is feature overlap to obtain the crown shape parameters. Details are not elaborated here.

[0033] The understory monitoring adjustment module is respectively connected to the forest feature acquisition module and the feature preprocessing module, and is used to select the adjustment method of the airborne lidar as determining the scanning frequency of the airborne lidar according to the growth coefficient and the distribution coefficient, or determining the laser scanning wavelength for the airborne lidar to scan the marked trees with feature overlap according to the crown shape parameters; Specifically, the present invention does not limit the specific structure of the underforest monitoring and adjustment module. Preferably, it can be a microprocessor, which is used to select the adjustment method of the airborne lidar, determine the scanning frequency of the airborne lidar, and determine the laser scanning wavelength for scanning each marked tree, which will not be elaborated here.

[0034] Among them, the underforest identification difference area is determined according to the terrain parameters.

[0035] Specifically, the underforest feature analysis module is used to determine the terrain parameters, where the underforest feature analysis module is used to construct a real-time terrain model of the monitoring sub-area based on the point cloud data; Specifically, the point cloud data acquisition points when constructing the real-time terrain model can be set by those skilled in the art according to the accuracy requirements of forest resource monitoring. The higher the accuracy requirements, the smaller the interval distance between the preset acquisition points in the feature monitoring sub-area. The value range of the interval distance can be [3, 10], and the interval unit is m. Preferably, the interval distance can be 5m.

[0036] Specifically, the terrain slope is determined based on the point cloud data to construct a real-time terrain model of the monitoring sub-area. The obtained point cloud data is subjected to data preprocessing, including data import, filtering processing, and denoising processing. The filtering algorithm is used to remove non-ground points such as vegetation, and at the same time, methods such as statistical filtering are used to remove noise points. Based on the filtered ground point cloud data, a suitable interpolation method such as inverse distance weighted interpolation is selected to generate a digital elevation model (DEM). Through the first-order difference algorithm, the elevation change rate of each grid unit in the DEM and its adjacent unit is calculated to obtain the terrain slope. The DEM and the slope map are combined, and a three-dimensional modeling software is used to construct a terrain model of the monitoring sub-area that can reflect the real-time terrain conditions, which will not be elaborated here.

[0037] The underforest feature analysis module is used to divide the real-time terrain model into several sub-model areas, divide the terrain model samples into several sample sub-model areas, and establish an association relationship between the sub-model areas and the sample sub-model areas; The underforest feature analysis module is used to calculate the morphological coincidence degree between the sub-model area and the sample sub-model area respectively, and determine the minimum value of the morphological coincidence degree as the terrain parameter of the monitoring sub-area; The terrain model samples are determined according to the historical monitoring data of the monitoring sub-area.

[0038] Specifically, the area covered by the terrain model can be divided into sub-model areas according to a rectangular grid. The number of sub-model areas is the product of the area of the monitoring sub-area and the division factor. The division factor can be set by those skilled in the art according to the accuracy requirements of forest resource monitoring. The higher the accuracy requirement, the smaller the division factor. The value range of the division factor can be [0.2, 0.4]. Preferably, the division factor can be 0.3. Exemplarily, a specific embodiment for determining the number of sub-model areas is given here. When the division factor is set to 0.3 and the area of the monitoring sub-area is obtained as 40,000 square meters, the number of sub-model areas is 12,000, and the area of each sub-model area is 3.33 square meters.

[0039] Specifically, the terrain model sample is determined based on the historical detection data obtained by the survey vehicle within a preset time period before airborne lidar monitoring. The preset time period can be set by those skilled in the art according to the accuracy requirements of forest resource monitoring. The higher the accuracy requirement, the smaller the change in the terrain itself, and the shorter the preset time period. The value range of the preset time period can be [5, 10], and the interval unit is days. Preferably, the preset time period can be 8 days.

[0040] Specifically, the morphological coincidence degree between the sub-model area and the sample sub-model area can be calculated based on the point cloud data. First, perform point cloud data matching, count the number of coincident points and the total number of points for each, and substitute them into the formula: the number of coincident points / (the total number of points in the sub-model area point cloud data + the total number of points in the sample sub-model area point cloud data - the number of coincident points) × 100% to calculate the morphological coincidence degree, which will not be elaborated here.

[0041] Please refer to Figure 2 as shown, which is the logic flowchart of the understory feature analysis module in the embodiment of the present invention for screening feature monitoring sub-areas. The understory feature analysis module is used to screen feature monitoring sub-areas, where if the terrain parameters of the monitoring sub-area meet the feature determination conditions, the understory feature analysis module will screen the monitoring sub-area as a feature monitoring sub-area; if the terrain parameters of the monitoring sub-area do not meet the feature determination conditions, the understory feature analysis module will not screen the monitoring sub-area; The feature determination condition is that the terrain parameter does not exceed a preset terrain threshold.

[0042] Specifically, the preset terrain threshold can be set by those skilled in the art according to the terrain fluctuation rate of the monitoring sub-area. The faster the terrain fluctuation rate, the smaller the set terrain threshold. The value range of the terrain threshold can be [0.6, 0.8]. Preferably, the terrain threshold can be 0.7.

[0043] Specifically, the present invention determines terrain parameters through the underforest feature analysis module based on the point cloud data of each monitoring sub-region to screen out the feature monitoring sub-regions. It can be understood that the laser pulses of airborne lidar are easily affected by the upper-layer vegetation, thus affecting the accuracy of the point cloud data. A real-time terrain model is constructed based on the point cloud data actually obtained by the lidar. The accuracy of the point cloud data currently obtained by the lidar is characterized by the coincidence degree between the real-time terrain model and the terrain model sample. The higher the coincidence degree, the higher the accuracy of the point cloud data, and the smaller the influence of the upper-layer vegetation on the airborne lidar detection in the current monitoring sub-region. The lower the coincidence degree, the lower the accuracy of the point cloud data, and the greater the influence of the upper-layer vegetation on the airborne lidar detection in the current monitoring sub-region. Screening out the regions with a large influence of the upper-layer vegetation on the airborne lidar detection, that is, the feature monitoring sub-regions, is helpful for taking more targeted data collection on these regions in subsequent monitoring. Furthermore, it realizes the rapid identification of the regions where the laser pulses have an impact during the monitoring process and improves the monitoring accuracy of forest resources.

[0044] Specifically, the feature preprocessing module is used to mark a number of tree points and determine the distribution coefficient. Among them, The feature preprocessing module determines the height values of a number of preset acquisition points in the geodetic coordinate system within the feature monitoring sub-region according to the point cloud data, and marks the acquisition points whose height values exceed the preset height threshold as the tree points; The feature preprocessing module calculates the minimum interval distance between any one feature tree point and the remaining tree points within the feature monitoring sub-region, and determines the minimum interval distance as the distribution factor of the feature tree point; Calculate the variance of the distribution factors of a number of feature tree points, and determine the variance of the distribution factors as the distribution coefficient of the feature monitoring sub-region.

[0045] Specifically, the setting of a number of preset acquisition points within the feature monitoring sub-region can be set by those skilled in the art according to the accuracy requirements of forest resource monitoring. The higher the accuracy requirements, the smaller the interval distance between the preset acquisition points within the feature monitoring sub-region. The value range of the interval distance can be [3, 10], and the interval unit is m. Preferably, the interval distance can be 5m.

[0046] Specifically, the interval distance is the horizontal distance between two tree points.

[0047] Specifically, the preset height threshold is the average height of the point cloud in the feature monitoring sub-region + (standard deviation × preset height factor). The standard deviation is the square root of the variance of the height values of the point cloud in the feature monitoring sub-region. The preset height factor can be set by those skilled in the art according to the size of the standard deviation. The smaller the standard deviation, the smaller the preset height factor. The value range of the height factor can be [1, 3]. Preferably, the height factor can be 2.

[0048] Exemplarily, a specific embodiment for determining the height threshold is given here. The average height of the point cloud in the feature monitoring sub-region is obtained as 10 m, the standard deviation is 2 m, and the set height factor is 2. Therefore, the height threshold is 14 m.

[0049] Specifically, the present invention determines the distribution coefficient by the feature preprocessing module according to the distribution of several tree points to select the growth coefficient of the feature monitoring sub-region determined according to the point cloud data, or determines whether there is feature overlap of the landmark tree according to the comparison between the position of the landmark tree and the understory recognition difference region, and obtains the crown shape parameters of the landmark tree with feature overlap. It can be understood that different tree distributions represent different forest structures and ecological environments. By selecting appropriate methods, the diversity of the forest can be monitored more carefully. Whether it is the overall characteristics of the uniformly distributed area or the special characteristics of the locally concentrated area, accurate data collection can be obtained, and then the actual situation of the forest can be reflected more comprehensively and truly. Selecting different parameters according to different tree distributions helps to adopt more targeted monitoring measures for different regions. For the area with uniformly distributed trees, a relatively unified monitoring strategy can be carried out, while for the area with uneven trees and local concentration, a more targeted monitoring method can be adopted, providing a more accurate basis for the scientific monitoring of forest resources. Furthermore, it realizes the rapid identification of the area affected by laser pulses during the monitoring process, adaptively adjusts the monitoring method of the lidar according to the morphological characteristics of the upper vegetation, and improves the monitoring accuracy of forest resources.

[0050] Specifically, if the distribution coefficient of the feature monitoring sub-region meets the uniform distribution determination condition, the feature preprocessing module selects the growth coefficients of several trees in the feature monitoring sub-region determined according to the point cloud data; Specifically, under the condition that the tree points are relatively evenly distributed, the growth coefficients of several trees in the feature monitoring sub-region are determined according to the point cloud data. It can be understood that when the tree distribution is uniform, the tree diameters obtained from the point cloud data can more accurately represent the size of the trees in the entire region. Based on these accurate diameter data, the degree of canopy shading of the understory vegetation can be evaluated more precisely. Because the uniform distribution of tree points makes the influence range of each tree relatively stable and predictable, unlike the uneven distribution where there are local dense shading or sparse unshaded situations, it can provide a more accurate basis for adjusting the lidar parameters to ensure that the laser can effectively penetrate the canopy gaps to reach the understory vegetation. Obtaining tree diameters under the condition of uniform tree point distribution and adjusting the lidar parameters accordingly helps to ensure that the understory vegetation data obtained in the entire monitoring region has high reliability. Furthermore, it realizes the rapid identification of the regions affected by laser pulses during the monitoring process, adaptively adjusts the monitoring method of the lidar according to the morphological characteristics of the upper-layer vegetation, and improves the monitoring accuracy of forest resources.

[0051] If the distribution coefficient of the feature monitoring sub-region does not meet the uniform distribution determination condition, the feature preprocessing module selects to determine whether there is feature overlap of the landmark tree according to the comparison between the position of the landmark tree and the understory recognition difference region, and obtains the canopy morphological parameters of the landmark tree with feature overlap. Specifically, under the condition that the tree points are relatively unevenly distributed, it is determined whether there is feature overlap of the landmark tree according to the comparison between the position of the landmark tree and the understory recognition difference region, and the canopy morphological parameters of the region with feature overlap are obtained. It can be understood that the landmark tree is the tree with relatively concentrated tree point distribution, and the understory recognition difference region is the region with a low coincidence degree between the terrain obtained by the lidar and the historical terrain. The landmark tree is located in the understory recognition difference region, indicating the deviation of the terrain obtained by the airborne lidar in real time, and there may be the influence of the upper-layer vegetation. Obtaining the canopy morphological parameters in these regions can evaluate the degree of canopy shading of the upper layer on the understory vegetation. The regular canopy shading is relatively uniform. Adjust the lidar parameters according to the actual shading situation to make the laser energy more reasonably distributed and improve the ability to penetrate the canopy to reach the understory vegetation. Furthermore, more reliable understory vegetation data can be obtained. Furthermore, it realizes the rapid identification of the regions affected by laser pulses during the monitoring process, adaptively adjusts the monitoring method of the lidar according to the morphological characteristics of the upper-layer vegetation, and improves the monitoring accuracy of forest resources.

[0052] The uniform distribution determination condition is that the distribution coefficient does not exceed a preset distribution threshold.

[0053] Specifically, the preset distribution threshold can be set by those skilled in the art according to the accuracy requirements of forest resource monitoring. The higher the accuracy requirements, the smaller the preset distribution threshold. The value range of the distribution threshold can be [5, 10]. Preferably, the distribution threshold can be 8.

[0054] Specifically, the feature preprocessing module is used to determine the growth coefficients of several trees in the feature monitoring sub-region, where the feature preprocessing module determines the diameters of the trees at different height levels according to the point cloud data, determines the average diameter as the growth factor of the trees, calculates the variance according to the growth factors of several trees in the feature monitoring sub-region, and determines the variance as the growth coefficient.

[0055] Specifically, the interval distance of the height levels is the product of the height value of the tree and the diameter value factor. The height value of the tree can be determined through the point cloud data. The diameter value factor can be set by those skilled in the art according to the accuracy requirements of forest resource monitoring. The higher the accuracy requirements, the smaller the diameter value factor is set. The value range of the diameter value factor can be [0.05, 0.2]. Preferably, the diameter value factor can be 0.1.

[0056] Specifically, to obtain the diameters of the trees at different height levels, the original point cloud data can be preprocessed to separate the ground points and the tree points, the single-tree point clouds are segmented using a clustering algorithm, horizontal sections are set at different height levels, the point clouds on the sections are extracted, and the point clouds on the section are fitted into a circle using fitting methods such as the least squares method. The diameter of the tree at this height level is calculated according to the parameters of the fitted circle, which will not be elaborated here.

[0057] Please refer to Figure 3 as shown, which is the logic flow chart of the feature preprocessing module of the embodiment of the present invention for determining whether there is feature overlap of the marked trees. The feature preprocessing module is used to determine whether there is feature overlap of the marked trees, where the feature preprocessing module marks the feature tree points whose distribution factors of the feature tree points do not exceed the preset distribution reference value as the marked trees, and determines the area corresponding to the terrain parameters in the feature monitoring sub-region as the underforest recognition difference area; If the marked tree is in the underforest recognition difference area, the feature preprocessing module determines that there is feature overlap of the marked tree and obtains the crown shape parameters; If the marked tree is not in the underforest recognition difference area, the feature preprocessing module determines that there is no feature overlap of the marked tree.

[0058] Specifically, the preset distribution reference value can be set by those skilled in the art according to the accuracy requirements of forest resource monitoring. The higher the accuracy requirement, the smaller the preset distribution reference value. The value range of the distribution reference value can be [2, 4]. Preferably, the distribution reference value can be 3.

[0059] Specifically, the feature preprocessing module is used to determine the crown shape parameters. The feature preprocessing module acquires the crown point clouds of each marked tree, divides the crown point clouds into several voxels, determines the number of point clouds in the voxels, calculates the variance based on the number of point clouds in several voxels, and determines the variance as the crown shape parameter.

[0060] Specifically, a voxel is the abbreviation of a volume element. A voxel is a small cubic unit divided in a three-dimensional data field, which represents a specific position and attribute in three-dimensional space. It is widely used in fields such as computer graphics, medical imaging, and geological exploration, and will not be elaborated here.

[0061] Specifically, the crown point clouds can be regularly divided in three-dimensional space according to a fixed voxel size. The setting of the voxel size can be determined by those skilled in the art according to the accuracy requirements of forest resource monitoring. The higher the accuracy requirement, the smaller the voxel size. Preferably, the voxel size can be 0.5m × 0.5m × 0.5m.

[0062] Specifically, to acquire the crown point clouds of each marked tree, the original point cloud data can be collected by lidar, preprocessed, the ground points can be removed by using a ground filtering algorithm to highlight the tree point clouds, and different trees can be segmented by using a clustering or region growing algorithm to obtain the crown point clouds, which will not be elaborated here.

[0063] Specifically, to determine the number of point clouds in the voxels, according to the position and size range of each voxel in three-dimensional space, all the point cloud data in a single voxel can be traversed, and a region judgment algorithm of the point cloud coordinates and the voxel boundary coordinates can be used to sequentially judge whether each point cloud is within the voxel space range. Each time a point cloud is determined to belong to the voxel, the number of point clouds in each voxel is counted, which will not be elaborated here.

[0064] Please refer to Figure 4 As shown, it is a logic flowchart of the adjustment method selected by the underforest monitoring adjustment module of the embodiment of the present invention for the airborne lidar. The underforest monitoring adjustment module is used to select the adjustment method of the airborne lidar. Among them, If the growth coefficients of several trees in the feature monitoring sub-region are obtained, the adjustment method selected by the underforest monitoring adjustment module for the airborne lidar is to determine the scanning frequency of the airborne lidar according to the growth coefficients and the distribution coefficients; If it is determined whether there is feature overlap in the marked trees and the crown shape parameters are determined, the understory monitoring and adjustment module selects the adjustment method of the airborne lidar as the laser scanning wavelength for scanning each marked tree according to the crown shape parameters.

[0065] Specifically, the scanning frequency is negatively correlated with the growth coefficient and the distribution coefficient respectively, and the laser scanning wavelength is positively correlated with the crown shape parameters.

[0066] Specifically, when the growth coefficient is 0.3 and the distribution coefficient is 2, the scanning frequency is 2800 Hz; when the growth coefficient is 0.7 and the distribution coefficient is 3.5, the scanning frequency is 2200 Hz; when the growth coefficient is 1.2 and the distribution coefficient is 4.8, the scanning frequency is 1500 Hz. The greater the growth coefficient and the greater the distribution coefficient, the smaller the scanning frequency. When the crown shape parameter is 50, the laser scanning wavelength is 808 nm; when the crown shape parameter is 90, the laser scanning wavelength is 1064 nm; when the crown shape parameter is 120, the laser scanning wavelength is 1550 nm. The greater the crown shape parameter, the smaller the laser scanning wavelength.

[0067] Specifically, it can be understood that regular crowns usually have a relatively uniform distribution of branches and leaves and a relatively simple geometric shape, and the scattering and occlusion laws of laser are relatively easy to predict. Lasers with shorter wavelengths have higher resolution and stronger reflection signals, which can capture the boundaries and detailed information of the crown more precisely, thereby more accurately identifying the spatial relationship between the crown and the understory vegetation, and helping to improve the recognition and classification accuracy of the understory vegetation.

[0068] So far, the technical solution of the present invention has been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the protection scope of the present invention.

[0069] The above are only the preferred embodiments of the present invention and are not used to limit the present invention; for those skilled in the art, the present invention can have various changes and modifications. Any modification, equivalent substitution, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A forest resource monitoring system based on airborne lidar data, characterized in that, Including: A forest feature acquisition module for controlling an airborne lidar to acquire point cloud data of a forest area to be measured; An understory feature analysis module connected to the forest feature acquisition module for dividing the forest area to be measured into several monitoring sub-areas, and determining terrain parameters based on the point cloud data of each monitoring sub-area to screen out feature monitoring sub-areas; A feature preprocessing module connected to the forest feature acquisition module and the understory feature analysis module respectively for marking several tree points based on the point cloud data of the feature monitoring sub-areas, determining a distribution coefficient according to the distribution of the tree points to select a growth coefficient for the feature monitoring sub-areas determined according to the point cloud data, or, determining whether there is feature overlap of the marked trees according to the comparison between the position of the marked trees and the understory recognition difference area, and obtaining the crown shape parameters of the marked trees with feature overlap; An understory monitoring adjustment module connected to the forest feature acquisition module and the feature preprocessing module respectively for selecting an adjustment method of the airborne lidar as determining the scanning frequency of the airborne lidar according to the growth coefficient and the distribution coefficient, or, determining the laser scanning wavelength for the airborne lidar to scan the marked trees with feature overlap according to the crown shape parameters; Wherein, the understory recognition difference area is determined according to the terrain parameters.

2. The forest resource monitoring system based on airborne lidar data according to claim 1, characterized in that, The understory feature analysis module is used to determine terrain parameters, wherein, The understory feature analysis module is used to construct a real-time terrain model of the monitoring sub-area according to the point cloud data; For dividing the real-time terrain model into several sub-model areas, dividing the terrain model samples into several sample sub-model areas, and establishing an association relationship between the sub-model areas and the sample sub-model areas; For calculating the morphological coincidence degree between the sub-model area and the sample sub-model area respectively, and determining the minimum morphological coincidence degree as the terrain parameter of the monitoring sub-area; The terrain model samples are determined according to the historical monitoring data of the monitoring sub-areas.

3. The forest resource monitoring system based on airborne lidar data according to claim 2, characterized in that, The understory feature analysis module is used to screen out feature monitoring sub-areas, wherein, If the terrain parameter of the monitoring sub-area meets the feature determination condition, the understory feature analysis module screens the monitoring sub-area as a feature monitoring sub-area; The feature determination condition is that the terrain parameter does not exceed a preset terrain threshold.

4. The forest resource monitoring system based on airborne lidar data according to claim 3, wherein, The feature preprocessing module is used to mark several tree points and determine the distribution coefficient, wherein, The feature preprocessing module determines the height values of several preset acquisition points in the geodetic coordinate system within the feature monitoring sub-area according to the point cloud data, and marks the acquisition points with height values exceeding the preset height threshold as the tree points; The feature preprocessing module calculates the minimum interval distance between any one feature tree point and the remaining tree points within the feature monitoring sub-area, and determines the minimum interval distance as the distribution factor of the feature tree point; Calculating the variance of the distribution factors of several feature tree points, and determining the variance of the distribution factors as the distribution coefficient of the feature monitoring sub-area.

5. The forest resource monitoring system based on airborne lidar data according to claim 4, wherein, If the distribution coefficient of the feature monitoring sub-area meets the uniform distribution determination condition, the feature preprocessing module selects the growth coefficients of several trees within the feature monitoring sub-area determined according to the point cloud data; If the distribution coefficient of the feature monitoring sub-region does not meet the uniform distribution determination condition, the feature preprocessing module determines whether there is feature overlap of the landmark trees based on the comparison between the location of the landmark trees and the understory recognition difference region, and obtains the crown shape parameters of the landmark trees with feature overlap; The uniform distribution determination condition is that the distribution coefficient does not exceed a preset distribution threshold.

6. The forest resource monitoring system based on airborne lidar data according to claim 5, characterized in that, The feature preprocessing module is used to determine the growth coefficients of several trees in the feature monitoring sub-region, where, The feature preprocessing module determines the diameters of the trees at different height levels according to the point cloud data, determines the average diameter as the growth factor of the tree, calculates the variance according to the growth factors of several trees in the feature monitoring sub-region, and determines the variance as the growth coefficient.

7. The forest resource monitoring system based on airborne lidar data according to claim 5, characterized in that, The feature preprocessing module is used to determine whether there is feature overlap of the landmark trees, where, The feature preprocessing module marks the feature tree points with a distribution factor not exceeding a preset distribution reference value as landmark trees, and determines the area corresponding to the terrain parameters in the feature monitoring sub-region as the understory recognition difference region; If the landmark tree is within the understory recognition difference region, the feature preprocessing module determines that the landmark tree has feature overlap and obtains the crown shape parameters.

8. The forest resource monitoring system based on airborne lidar data according to claim 7, wherein, The feature preprocessing module is used to determine the crown shape parameters, where, The feature preprocessing module obtains the crown point cloud of each landmark tree, divides the crown point cloud into several voxels, determines the number of point clouds in the voxels, calculates the variance according to the number of point clouds in several voxels, and determines the variance as the crown shape parameter.

9. The forest resource monitoring system based on airborne lidar data according to claim 6 or 8, characterized in that The understory monitoring adjustment module is used to select the adjustment method of the airborne lidar, where, If the growth coefficients of several trees in the feature monitoring sub-region are obtained, the understory monitoring adjustment module selects the adjustment method of the airborne lidar as determining the scanning frequency of the airborne lidar according to the growth coefficient and the distribution coefficient; If it is determined whether there is feature overlap of the landmark trees and the crown shape parameters are determined, the understory monitoring adjustment module selects the adjustment method of the airborne lidar as determining the laser scanning wavelength for scanning each landmark tree according to the crown shape parameters.

10. The forest resource monitoring system based on airborne lidar data according to claim 9, characterized in that, The scanning frequency is negatively correlated with the growth coefficient and the distribution coefficient respectively, and the laser scanning wavelength is positively correlated with the crown shape parameter.

Images