Karst forest canopy density estimation method and system suitable for planting under Chinese herbal medicine forest

Abstract

The invention discloses a karst forest canopy density estimation method suitable for planting under a traditional Chinese medicinal material forest. The method comprises the following steps: acquiring three-dimensional laser point cloud data and surface spectrum data of a target area, and a canopy density true value of a ground verification quadrat in the target area; extracting ground points and non-ground points, extracting vegetation points, and calculating initial canopy density of each grid unit and ground verification quadrat; constructing a correction model and training, wherein a canopy density measurement error is a difference value between a canopy density truth value and the initial canopy density; and obtaining the corrected canopy density corresponding to each grid unit based on the correction model. According to the method, the three-dimensional laser point cloud data of the LiDAR and the multispectral surface spectral data are fused, and the correction model for karst landform features is introduced for the first time, so that the interference of the gradient and the rock exposure rate on canopy density estimation is fundamentally solved, and the precision and the reliability of an estimation result are far better than those of an existing method.

Description

Technical Field

[0001] This invention relates to the fields of forest resource survey and precision forestry technology, and more specifically to a method and system for estimating canopy closure in karst forests suitable for understory planting of medicinal herbs. More specifically, it addresses the problem of low accuracy in traditional canopy closure measurement methods due to complex terrain and high rock exposure in karst landforms, and is particularly suitable for guiding the scientific planting of understory economic crops (such as medicinal herbs) that have specific light requirements. Technical Background

[0002] Forest canopy density, defined as the ratio of the vertical projection area of ​​the forest canopy to the total forest area, is one of the core indicators for evaluating forest structure, health status, and the forest's ecological environment. It directly determines the microclimate conditions under the forest canopy, such as light intensity, temperature, humidity, and soil moisture, thus profoundly affecting the growth, development, yield, and even the accumulation of medicinal components in the understory vegetation.

[0003] In recent years, with the rise of understory economy, developing understory medicinal herb cultivation using forest resources has become an important way to promote local economic development and ecological restoration. Different medicinal herbs have extremely demanding and varied requirements for light environment. For example, research shows that the suitable canopy closure for Dendrobium officinale under forest cover is 0.4-0.6, while other medicinal herbs may require higher or lower canopy closure conditions to achieve optimal growth and active ingredient content. Therefore, accurately and efficiently obtaining large-scale, high-precision spatial distribution maps of forest canopy closure is a prerequisite for the scientific layout and industrial development of understory medicinal herbs.

[0004] However, existing methods for measuring canopy closure have the following significant drawbacks when applied to the cultivation of medicinal herbs under forest canopies in karst areas: 1) Subjectivity and inefficiency of traditional ground survey methods: Traditional methods such as visual estimation, sampling points, and transects are still widely used. Visual estimation is simple to operate, but its results heavily rely on the experience of the surveyors, are highly subjective, and show significant differences between different surveyors, making it difficult to meet the requirements of precision agriculture. More precise methods, such as canopy projection, are time-consuming and labor-intensive, and not suitable for large-scale rapid surveys. 2) The accuracy bottleneck of traditional remote sensing estimation methods in karst landforms: Estimating canopy closure using satellite or UAV remote sensing images has been a research hotspot in recent years. These methods usually establish models based on the statistical relationship between vegetation indices (such as NDVI) and canopy closure. However, karst landforms are characterized by rapid topographic changes, fragmented surfaces, steep slopes, and high rock exposure rates. These characteristics significantly interfere with remote sensing estimation: ① Topographic shading effect: In steep mountainous areas, topographic shading severely affects the spectral information of optical images, leading to distorted vegetation indices and thus underestimating or overestimating canopy closure. ② Rock-vegetation mixed pixel problem: Karst regions are characterized by rugged terrain with interspersed rocks and vegetation, often containing multiple land features within a single pixel in remote sensing images. High rock exposure significantly reduces vegetation indices, causing models to misclassify areas as having low canopy closure, when in reality the canopy may still be intact. ③ Limitations of two-dimensional information: Traditional optical remote sensing acquires two-dimensional information about the Earth's surface and cannot penetrate the canopy to obtain its three-dimensional structure. Therefore, it is difficult to distinguish between tall trees and low shrubs and grasses, and it cannot directly characterize the vertical structure of the canopy, which limits the physical meaning and accuracy of canopy closure estimation. 3) Lack of targeted error correction mechanisms: Although existing studies have recognized the impact of topography on remote sensing estimation, few studies have been able to quantify the specific impact of slope, aspect and rock exposure rate unique to karst areas on the measurement error of canopy closure, and even fewer have proposed a systematic error correction model. This results in poor applicability of existing models in karst areas and difficulty in guaranteeing accuracy.

[0005] In summary, there is an urgent need for a new method that can overcome the interference of complex karst terrain and achieve high-precision, high-efficiency, and highly objective canopy closure estimation to meet the pressing need for precise cultivation of medicinal herbs under forest canopies. The technical solution proposed in this patent application is precisely designed to solve this technical challenge. Summary of the Invention

[0006] The purpose of this invention is to address the problems of low accuracy and poor reliability in estimating forest canopy closure in karst regions due to complex terrain and high rock exposure. Its core objective is to provide a method and system for estimating canopy closure in karst forests suitable for understory cultivation of medicinal herbs. This system can integrate multi-source remote sensing data and incorporate a built-in geomorphological feature error correction model for high-precision canopy closure estimation, providing reliable data support for the scientific cultivation of medicinal herbs understory in karst regions.

[0007] To achieve the above objectives, the present invention proposes the following technical solution:

[0008] A method for estimating the canopy closure of karst forests suitable for understory planting of Chinese medicinal herbs includes the following steps:

[0009] Step 1: Obtain the 3D laser point cloud data, surface spectral data, and ground verification plot canopy closure true values ​​of the target area;

[0010] Step 2: Extract ground points and non-ground points based on 3D laser point cloud data, extract vegetation points based on non-ground points, divide grid cells based on 3D laser point cloud data, and calculate the preliminary canopy closure of each grid cell and ground verification plot.

[0011] Step 3: Construct and train a calibration model with the canopy closure measurement error of the ground verification plot as the dependent variable and the slope and rock exposure rate corresponding to the ground verification plot as the independent variables. The canopy closure measurement error is the difference between the true canopy closure value and the preliminary canopy closure value.

[0012] Step 4: Input the slope and rock exposure rate corresponding to each grid cell in the target area into the trained calibration model to obtain the canopy closure measurement error corresponding to each grid cell, and further obtain the calibrated canopy closure corresponding to each grid cell.

[0013] As mentioned above, step 1 also includes coordinate registration of the three-dimensional laser point cloud data and the surface spectral data.

[0014] As described above, the slope is obtained based on the following steps:

[0015] A high-precision digital terrain model is generated using ground points. A canopy height model is calculated using vegetation points and the digital terrain model. The slope of each grid cell and ground verification plot is calculated based on the high-precision digital terrain model.

[0016] As mentioned above, the calibration model is a multivariate nonlinear regression model, a random forest regression model, or a gradient boosting regression tree model.

[0017] As mentioned above, the correction for canopy closure is based on the following formula:

[0018] CC corrected =CC initial+f(S,R)

[0019] Among them, CC corrected To correct for canopy closure, CC initial , where f() is the initial canopy closure, S is the slope, and R is the rock exposure rate.

[0020] As described above, the multivariate nonlinear regression model is based on the following formula

[0021] Error CC =α×S^β+γ×R^δ+ε;

[0022] Where α, β, γ, and δ are fitting parameters, ε is the residual term, and ^ represents the power sign. CC This represents the measurement error of canopy closure.

[0023] The method for estimating the canopy closure of karst forests suitable for understory planting of Chinese medicinal herbs also includes the following steps:

[0024] Step 5: Automatically divide the grid cells based on the corrected canopy closure, and automatically divide the suitable areas for understory medicinal herbs in the target area.

[0025] A karst forest canopy closure estimation system suitable for understory planting of Chinese medicinal herbs includes:

[0026] Unmanned aerial vehicle (UAV) platforms for carrying LiDAR sensors and multispectral or hyperspectral cameras;

[0027] Ground verification equipment is used to collect the true values ​​of canopy closure for each ground verification quadrat.

[0028] The data processing unit is used to perform steps 1 to 5 of the above-mentioned method for estimating the canopy closure of karst forests suitable for planting Chinese medicinal herbs under forest canopies;

[0029] The output unit is used to output the corrected canopy closure distribution map.

[0030] Compared with the prior art, the technical solution of the present invention has the following significant advantages:

[0031] ① High accuracy and reliability: By fusing LiDAR's three-dimensional laser point cloud data with multispectral surface spectral data, and pioneering the introduction of a correction model for karst landform characteristics, the interference of slope and rock exposure rate on canopy closure estimation is fundamentally solved, making the accuracy and reliability of the estimation results far exceed those of existing methods.

[0032] ② Strong objectivity and repeatability: The entire process is based on sensor data and algorithm models, eliminating the subjective factors of traditional manual surveys, ensuring the objectivity, consistency and repeatability of the results, and meeting the standardized requirements of scientific research and industrial production.

[0033] ③ Low cost and high efficiency: Using drones as a platform, surveys of large areas of mountains and forests can be completed in a short time, which is more than 100 times more efficient than traditional ground methods, greatly reducing manpower and time costs.

[0034] ④ Precise guidance for understory planting: The generated sub-meter-level high-precision canopy closure distribution map can clearly delineate microhabitats under different light conditions. For example, areas with canopy closure between 0.4 and 0.6 can be precisely delineated specifically for planting Dendrobium; highly closed areas with canopy closure greater than 0.7 can be delineated for planting more shade-tolerant medicinal herbs. This "map-based" precision planting will greatly improve the survival rate, yield, and quality of medicinal herbs, promoting the transformation of the understory economy in karst regions towards refinement and scientific management. Attached Figure Description

[0035] Figure 1 This is a flowchart of the present invention;

[0036] Figure 2 This is a comparison chart of canopy closure accuracy before and after correction.

[0037] Figure 3 A comparison chart of the accuracy of the corrected canopy closure;

[0038] Figure 4 This is a high-precision corrected canopy closure distribution map after correction. Detailed Implementation

[0039] To facilitate understanding and implementation of the present invention by those skilled in the art, the present invention will be further described in detail below with reference to embodiments. The embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0040] Example 1

[0041] like Figure 1 As shown, the method for estimating the canopy closure of karst forests suitable for understory planting of Chinese medicinal herbs includes the following steps:

[0042] Step 1: Receive multi-source heterogeneous data, which includes three-dimensional laser point cloud data of the target area, surface spectral data, and the true canopy closure (CCtrue) of the ground verification plots in the target area.

[0043] 3D laser point cloud data acquisition: Using a drone equipped with a high-frequency lidar scanner (UAV-LiDAR), flight operations are conducted over the target area to acquire high-density 3D laser point cloud data. The UAV-LiDAR can penetrate part of the canopy, directly acquiring the 3D coordinate information of the canopy and the ground, providing a foundation for accurately separating the ground from vegetation and constructing digital terrain models (DTM) and canopy height models (CHM).

[0044] Surface spectral data acquisition: High-resolution surface spectral data of the target area are simultaneously acquired using a multispectral or hyperspectral camera on the same UAV platform or via a closely following flight. Surface spectral data is used to accurately identify land cover types, particularly distinguishing between vegetation, bare soil, and exposed rock.

[0045] Coordinate registration is performed between 3D laser point cloud data and surface spectral data.

[0046] True canopy closure data acquisition: Ground validation quadrats are deployed within the target area to perform simultaneous multi-parameter validation. This is crucial for model training and accuracy evaluation.

[0047] Step 2: Data Preprocessing and Feature Extraction

[0048] Point cloud denoising and classification: The original 3D laser point cloud data is denoised, and ground point classification algorithms (such as cloth simulation filtering or progressive TIN encryption algorithm) are used to accurately classify the point cloud into ground points and non-ground points (vegetation points, building points, etc.).

[0049] Vegetation point extraction based on non-ground points: Vegetation points are selected from non-ground points based on the geometric features (such as height, density, and shape) or spectral characteristics of vegetation. Considering the limitations of 3D point cloud data, relying solely on geometric features may be insufficient to accurately distinguish vegetation from other non-ground targets. Therefore, in practice, it may be necessary to combine more data or features, such as spectral information and temporal dynamic changes, to improve the accuracy of vegetation point extraction.

[0050] Topography and Canopy Model Generation: A high-precision digital terrain model (DTM) is generated using ground points; a canopy height model (CHM) is calculated using vegetation points and the digital terrain model (DTM), which reflects the true height of the trees.

[0051] Preliminary Canopy Closure Estimation (CCinitial): The classified 3D laser point cloud data is divided into grid cells, and the preliminary canopy closure of each grid cell and the ground verification plot is calculated using the "point cloud ratio method." Specifically, within each grid cell, the number of vegetation points above a certain threshold (e.g., 2 meters to exclude the influence of shrubs and grasses) is divided by the total number of point cloud points within that grid cell to obtain the preliminary canopy closure corresponding to each grid cell and the ground verification plot. The physical meaning of the above steps is clear and superior to traditional 2D imagery methods.

[0052] Quantification of karst landform disturbance factors: The slope factor is quantified based on a high-precision digital terrain model (DTM) to calculate the slope (S) of each grid cell and ground validation plot. The rock exposure rate factor is calculated accurately for each grid cell and ground validation plot using synchronously acquired surface spectral data through supervised classification or spectral index methods (such as constructing a rock index).

[0053] Step 3: Core Innovation: Construction and Training of Karst Landform Error Correction Model

[0054] Error Source Analysis: The core assumption of this invention is that the initial canopy closure (CCinitial) of each grid cell in the karst region is the source of error. CC The main causes are slope (S) and rock exposure (R). For example, in areas with steep slopes, the angle and point density distribution of UAV LiDAR scanning may change, potentially leading to errors; in areas with high rock exposure, although there may be no vegetation under the forest canopy, the canopy cover above may still exist, which can easily cause confusion using traditional methods.

[0055] Model construction: using true canopy closure values ​​(CC) obtained from ground validation quadrats. ture (Measured using high-precision instruments such as LAI-2200C), and compared with the preliminary canopy closure (CC) of the ground verification plots estimated by UAV. initial By comparing these data, the measurement error of canopy closure for each ground validation plot was obtained. CC =CC true -CC initial Then, using the canopy closure measurement error of the ground validation plots as the dependent variable and the slope (S) and rock exposure rate (R) of the corresponding ground validation plots as independent variables, a multivariate nonlinear regression model or machine learning model (such as a random forest regression model or a gradient boosting regression tree model) is constructed as a calibration model. The calibration model is trained using the slope (S), rock exposure rate (R), and corresponding initial canopy closure (CCinitial) of each ground validation plot. Here, the calibration model expression is inferred as: Error CC = f(S,R), where f() is the correction model, and thus the final corrected canopy closure (CC) is obtained. corrected ):CC corrected =CC initial +f(S,R).

[0056] As a preferred option, the calibration model employs a multivariate nonlinear regression model, which is based on the following formula:

[0057] Error CC =α×S^β+γ×R^δ+ε;

[0058] Where α, β, γ, and δ are fitting parameters, ε is the residual term, and ^ is the power sign.

[0059] Step 4: Input the slope (S) and rock exposure rate (R) corresponding to each grid cell in the target area into the trained calibration model to obtain the canopy closure measurement error for each grid cell. CC Based on the initial canopy closure (CCinitial) and canopy closure measurement error (Error) of each grid cell, CC The corrected canopy closure for each grid cell is calculated, and the trained correction model is applied to the entire target area. The initial canopy closure value of each grid cell is corrected one by one, and finally a high-precision, high-resolution map of the corrected canopy closure distribution of the karst forest is generated.

[0060] Accuracy Verification and Evaluation: The accuracy of the corrected canopy closure distribution map is evaluated using cross-validation methods or independent validation sample sets. The evaluation follows international standards, such as the principles regarding the accuracy (correctness and precision) of measurement methods in ISO 5725, systematically assessing the repeatability and reproducibility of the model. The root mean square error (RMSE), mean absolute error (MAE), and coefficient of determination (R²) are calculated. 2 The accuracy of the method of this invention is quantified by comparing statistical indicators such as (i) and uncorrected results with those of traditional remote sensing methods.

[0061] Step 5: Further, the grid cells are automatically divided based on the corrected canopy closure, automatically dividing the suitable areas for understory medicinal herbs in the target area, and generating a planting suggestion layer based on the suitable areas for understory medicinal herbs. Specifically, when the corrected canopy closure is greater than 0.4 and less than or equal to 0.6, the corresponding grid cell is used as the planting suggestion grid cell for Dendrobium, and a planting suggestion layer for Dendrobium is generated based on each Dendrobium planting suggestion grid cell; when the corrected canopy closure is greater than or equal to 0.7, the corresponding grid cell is used as the planting suggestion grid cell for shade-tolerant medicinal herbs, and a planting suggestion layer for shade-tolerant medicinal herbs is generated based on each shade-tolerant medicinal herb planting suggestion grid cell.

[0062] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods.

[0063] Example 2:

[0064] 1) Data acquisition and equipment

[0065] Target area: Select a typical karst forest area with peaks and depressions, covering an area of ​​about 5 square kilometers, with large topographic relief and moderate degree of rocky desertification.

[0066] Drone and Sensors: The drone is a DJI M600 Pro, equipped with a Velodyne VLP-16LiDAR sensor and a MicaSense RedEdge-MX multispectral camera.

[0067] Flight parameters: flight altitude 100 meters, forward overlap rate 80%, lateral overlap rate 70%, ensuring that the point cloud density of the 3D laser point cloud data is not less than 200 points / square meter, and the resolution of the ground surface spectral data is better than 10 centimeters.

[0068] 2) Ground verification scheme design

[0069] Plot layout: Within the target area, 60 20m×20m ground plots were laid out using stratified random sampling based on two gradients: slope (less than 15°, greater than or equal to 15° and less than 30°, greater than or equal to 30°) and rock exposure rate (less than 20%, greater than or equal to 20% and less than 50%, greater than or equal to 50%).

[0070] True canopy closure measurement: Within each ground validation plot, a LAI-2200C plant canopy analyzer was used to perform measurements at the center and four corner points (or a nine-square grid) of the ground validation plot using the standard "one up, four down" or "one up, nine down" method to obtain high-precision true canopy closure values ​​(CCtrue). All measurements were performed under diffused light conditions (early morning, evening, or cloudy days) to eliminate the influence of direct sunlight. Operation was strictly performed in accordance with the instrument operating instructions to ensure data quality.

[0071] Simultaneous verification of light environment: While measuring the true value of canopy closure, the photosynthetically active radiation intensity at a set height (simulating the growth height of medicinal herbs) under the forest canopy is measured simultaneously using equipment equipped with a photosynthetically active radiation (PAR) sensor. This establishes a direct relationship between the true value of canopy closure and the actual available light energy, providing more direct guidance for medicinal herb cultivation.

[0072] Data quality control: The entire ground validation process references the concept of intermediate accuracy in ISO 5725-3 standard. Two operators repeatedly measure some ground validation plots at different times to evaluate and ensure the stability and reliability of the measurement results.

[0073] 3) Data processing and modeling

[0074] In step 2, the point cloud of the 3D laser point cloud data is classified to generate a digital terrain model (DTM) and a canopy height model (CHM) with a resolution of 0.5 meters.

[0075] In step 3, 45 samples are randomly selected from 60 samples as the training set and 15 samples are selected as the validation set. The calibration model adopts the random forest regression model.

[0076] Implementation Example 1 is carried out based on the above conditions.

[0077] 4) Results and Analysis

[0078] ① Accuracy Comparison: Using 15 validation samples, the initial canopy closure (CCinitial) and the corrected canopy closure (CCcorrected) were compared with the true canopy closure (CCtrue). The results showed that the RMSE of the initial canopy closure (CCinitial) was 0.18, and R... 2 The RMSE for corrected canopy closure (CC) was 0.65; while the RMSE for CC-corrected decreased to 0.07. 2 Increased to 0.92 ( Figures 2-3 This indicates that the correction model of the present invention is highly effective.

[0079] Suitable zoning map generation: based on the corrected high-precision corrected canopy closure distribution map ( Figure 4 The study area was divided into three zones: Optimal Dendrobium Zone: Corrected canopy closure greater than 0.4 and less than or equal to 0.6, marked in green. Secondary Optimal Dendrobium Zone: Corrected canopy closure greater than 0.3 and less than or equal to 0.4, or greater than 0.6 and less than or equal to 0.7, marked in yellow. Inappropriate Dendrobium Zone: Canopy closure less than or equal to 0.3, or greater than 0.7, marked in red.

[0080] The generated corrected canopy closure distribution map can be directly delivered to local forest farmers or enterprises to guide the precise planting, cultivation management and yield prediction of Dendrobium, realizing a closed loop from data to decision-making.

[0081] Example 3:

[0082] A system for high-precision estimation of forest canopy density in karst landforms includes:

[0083] Unmanned aerial vehicle (UAV) platforms for carrying LiDAR sensors and multispectral or hyperspectral cameras;

[0084] Ground verification equipment is used to collect the true values ​​of canopy closure for each ground verification quadrat.

[0085] The data processing unit is used to execute steps 1 to 5 of Embodiment 1;

[0086] The output unit is used to output the corrected canopy closure distribution map.

[0087] Example 4:

[0088] A computer device includes a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0089] Example 5:

[0090] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps in the above-described method embodiments.

[0091] Example 6:

[0092] A computer program product includes a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0093] It should be noted that the embodiments described in this invention are merely illustrative examples. Those skilled in the art to which this invention pertains can make various modifications or additions to the described embodiments or use similar methods to substitute them, without departing from the spirit of this invention or exceeding the scope defined by the appended claims.

Claims

1. A method for estimating canopy closure in karst forests suitable for understory planting of medicinal herbs, characterized in that, Includes the following steps: Step 1: Obtain the 3D laser point cloud data, surface spectral data, and ground verification plots true canopy closure values ​​of the target area; Step 2: Extract ground points and non-ground points based on 3D laser point cloud data, extract vegetation points based on non-ground points, divide grid cells based on 3D laser point cloud data, and calculate the preliminary canopy closure of each grid cell and ground verification plot. Step 3: Construct and train a calibration model with the canopy closure measurement error of the ground verification plot as the dependent variable and the slope and rock exposure rate corresponding to the ground verification plot as the independent variables. The canopy closure measurement error is the difference between the true canopy closure value and the preliminary canopy closure value. Step 4: Input the slope and rock exposure rate corresponding to each grid cell in the target area into the trained calibration model to obtain the canopy closure measurement error corresponding to each grid cell, and further obtain the calibrated canopy closure corresponding to each grid cell.

2. The method for estimating canopy closure in karst forests suitable for understory planting of medicinal herbs according to claim 1, characterized in that, Step 1 also includes coordinate registration of the three-dimensional laser point cloud data and the surface spectral data.

3. The method for estimating canopy closure in karst forests suitable for understory planting of medicinal herbs according to claim 1, characterized in that, The slope is obtained based on the following steps: A high-precision digital terrain model is generated using ground points. A canopy height model is calculated using vegetation points and the digital terrain model. The slope of each grid cell and ground verification plot is calculated based on the high-precision digital terrain model.

4. The method for estimating canopy closure in karst forests suitable for understory planting of medicinal herbs according to claim 1, characterized in that, The correction model is a multivariate nonlinear regression model, a random forest regression model, or a gradient boosting regression tree model.

5. The method for estimating canopy closure in karst forests suitable for understory planting of medicinal herbs according to claim 1, characterized in that, The corrected canopy closure is based on the following formula: CC corrected =CC initial +f(S,R) Among them, CC corrected To correct for canopy closure, CC initial , where f() is the initial canopy closure, S is the slope, and R is the rock exposure rate.

6. The method for estimating canopy closure in karst forests suitable for understory planting of medicinal herbs according to claim 4, characterized in that, The multivariate nonlinear regression model is based on the following formula Error CC =α×S^β+γ×R^δ+ε; Where α, β, γ, and δ are fitting parameters, ε is the residual term, and ^ represents the power sign. CC This represents the measurement error of canopy closure.

7. The method for estimating canopy closure in karst forests suitable for understory planting of medicinal herbs according to claim 1, characterized in that, It also includes the following steps: Step 5: Automatically divide the grid cells based on the corrected canopy closure, and automatically divide the suitable areas for understory medicinal herbs in the target area.

8. A karst forest canopy closure estimation system suitable for understory planting of Chinese medicinal herbs, characterized in that, include: Unmanned aerial vehicle (UAV) platforms for carrying LiDAR sensors and multispectral or hyperspectral cameras; Ground verification equipment is used to collect the true values ​​of canopy closure for each ground verification quadrat. The data processing unit is used to perform steps 1 to 5 of the method for estimating the canopy closure of karst forests suitable for planting medicinal herbs under forests as described in claim 7. The output unit is used to output the corrected canopy closure distribution map.

Images