CN110148116A - A kind of forest biomass evaluation method and its system - Google Patents
A kind of forest biomass evaluation method and its system Download PDFInfo
- Publication number
- CN110148116A CN110148116A CN201910295160.6A CN201910295160A CN110148116A CN 110148116 A CN110148116 A CN 110148116A CN 201910295160 A CN201910295160 A CN 201910295160A CN 110148116 A CN110148116 A CN 110148116A
- Authority
- CN
- China
- Prior art keywords
- data
- point cloud
- biomass
- forest
- lidar point
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 239000002028 Biomass Substances 0.000 title claims abstract description 107
- 238000011156 evaluation Methods 0.000 title abstract 2
- 238000000034 method Methods 0.000 claims abstract description 48
- 238000005070 sampling Methods 0.000 claims abstract description 9
- 238000012937 correction Methods 0.000 claims description 23
- 238000010586 diagram Methods 0.000 claims description 19
- 230000011218 segmentation Effects 0.000 claims description 19
- 238000010606 normalization Methods 0.000 claims description 18
- 238000012545 processing Methods 0.000 claims description 18
- 230000005855 radiation Effects 0.000 claims description 12
- 238000003064 k means clustering Methods 0.000 claims description 4
- 238000002310 reflectometry Methods 0.000 claims description 4
- 230000003595 spectral effect Effects 0.000 abstract description 4
- 210000005252 bulbus oculi Anatomy 0.000 abstract 1
- 238000009499 grossing Methods 0.000 description 8
- 238000012360 testing method Methods 0.000 description 6
- 238000005259 measurement Methods 0.000 description 4
- 238000011835 investigation Methods 0.000 description 3
- 210000000481 breast Anatomy 0.000 description 2
- 238000001035 drying Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 230000001788 irregular Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000000926 separation method Methods 0.000 description 2
- 239000002023 wood Substances 0.000 description 2
- 241000196324 Embryophyta Species 0.000 description 1
- 244000166124 Eucalyptus globulus Species 0.000 description 1
- 230000004931 aggregating effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 239000011362 coarse particle Substances 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000012217 deletion Methods 0.000 description 1
- 230000037430 deletion Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000011065 in-situ storage Methods 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 230000035764 nutrition Effects 0.000 description 1
- 235000016709 nutrition Nutrition 0.000 description 1
- 238000001228 spectrum Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10028—Range image; Depth image; 3D point clouds
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10032—Satellite or aerial image; Remote sensing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10032—Satellite or aerial image; Remote sensing
- G06T2207/10036—Multispectral image; Hyperspectral image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30181—Earth observation
- G06T2207/30188—Vegetation; Agriculture
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Quality & Reliability (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a kind of forest biomass evaluation method and its systems.The method includes the steps: the forest parameters for obtaining region to be measured calculate the forest biomass of sampling point；LiDAR point cloud data are obtained, and the LiDAR point cloud data are separated, digital elevation model is generated according to the ground point；The LiDAR point cloud data are normalized, normalized LiDAR point cloud data are obtained；The raster data that the normalized LiDAR point cloud data are switched to predetermined resolution, obtains trees height grid data；The vegetation index distributed data in the region is calculated according to the multispectral data；The forest biomass and trees height grid data and vegetation index distributed data are estimated to the forest biomass in region to be measured using biomass regression formula.The present invention combines the spectral information of unmanned plane multispectral data and the forest three-dimensional structure information of unmanned plane LiDAR data, carries out inverting estimation using the truthful data of eyeball, substantially increases the precision of inverting.
Description
Technical Field
The invention relates to the technical field of remote sensing inversion, in particular to a forest biomass estimation method and a forest biomass estimation system.
Background
The traditional forest resource survey mainly acquires related data by field measurement, and the method has the defects of time and labor consumption, strong subjectivity, time and labor consumption for searching, great destructiveness, limitation to estimation of the biomass of the existing biomass and forest stand biomass in a small range, and unsuitability for large-area forest survey. And the appearance of remote sensing technology provides possibility for the investigation of forest resources in a large range. However, the commonly used satellite optical remote sensing image can only provide spectral information of the forest, important three-dimensional structure information is difficult to obtain, and the resolution of part of the image is low and is easily influenced by weather, so that the problem of low precision in the process of obtaining forest biomass exists.
The emerging LiDAR technology can provide forest vertical structure information which cannot be reflected by passive optical remote sensing while acquiring forest horizontal distribution information, and has better advantages in reflecting ground biomass. However, because LiDAR cannot provide information such as forest spectrum and vegetation index, there is still a certain limitation on inverting forest biomass.
Accordingly, the prior art is yet to be improved and developed.
Disclosure of Invention
The present invention provides a forest biomass estimation method and a system thereof, aiming at solving the above-mentioned drawbacks of the prior art, and aims to solve the problem that the forest biomass estimation method in the prior art has certain limitations.
The technical scheme adopted by the invention for solving the technical problems is as follows:
a forest biomass estimation method comprises the following steps:
calculating the forest biomass of the sampling point by acquiring forest parameters of the area to be measured;
the method comprises the steps of obtaining LiDAR point cloud data of a region to be measured, and separating the LiDAR point cloud data to obtain ground points; generating a digital elevation model according to the ground points;
performing normalization processing on the LiDAR point cloud data by adopting the digital elevation model to obtain normalized LiDAR point cloud data;
converting the normalized LiDAR point cloud data into raster data with a preset resolution ratio to obtain tree height raster data;
acquiring multispectral data of an area to be measured, and calculating vegetation index distribution data of the area according to the multispectral data;
and estimating the forest biomass of the area to be measured by using the forest biomass, the tree height grid data and the vegetation index distribution data through a biomass regression formula.
The forest biomass estimation method comprises the following steps of obtaining multispectral data of a region to be measured, and calculating vegetation index distribution data of the region according to the multispectral data, and specifically comprises the following steps:
according to the aerial photography geographic coordinate control points, carrying out image splicing on unmanned aerial vehicle data images of the area to be tested;
and carrying out radiation correction on the spliced unmanned aerial vehicle data image, and calculating a normalized vegetation index of vegetation in the region to be detected.
The forest biomass estimation method comprises the following steps of converting the normalized LiDAR point cloud data into raster data with a preset resolution ratio to obtain tree height raster data, and specifically comprises the following steps:
carrying out normalization processing on the LiDAR point cloud data to obtain the absolute height of the tree to the ground;
carrying out horizontal layering on the normalized LiDAR point cloud data according to a preset layering setting; the preset layering setting means that layering is carried out at certain layering intervals from a specified position;
carrying out point cloud clustering on each layer obtained by layering by using a K-Means clustering method to obtain a cluster;
setting cluster polygons, and stacking the cluster polygons to obtain a stack diagram;
identifying a local maximum value in the stacking graph by adopting a window with a fixed size to obtain a single tree parameter;
and taking the single tree parameters as seed points, performing point cloud single tree segmentation on the seed points by adopting a PCS algorithm, and editing the segmentation result to obtain tree height grid data.
The forest biomass estimation method comprises the following steps of:
wherein AGB is ground biomass, H is tree height, and NDVI is a normalized vegetation index.
The forest biomass estimation method comprises the following steps of carrying out radiation correction on the spliced unmanned aerial vehicle data image, and calculating a normalized vegetation index of vegetation in a region to be measured, wherein the radiation correction comprises the following steps: green band correction, near infrared band correction, red edge band correction, and red band correction.
The forest biomass estimation method comprises the following steps of:
wherein,andrespectively representing the reflectivity of the near infrared band and the red light band.
The forest biomass estimation method, wherein the predetermined resolution is a resolution of 0.3-0.6 meters.
The forest biomass estimation method is characterized in that the designated position is 0.5-0.7 m, and the layering interval is 1 m.
The forest biomass estimation method, wherein the fixed size is 1.0-2.0 meters.
A surface biomass estimation system, comprising: a processor, and a memory coupled to the processor;
the memory stores a terrestrial biomass estimation program that when executed by the processor performs the steps of:
acquiring forest biomass of sampling points in an area to be detected;
the method comprises the steps of obtaining LiDAR point cloud data of a region to be measured, and separating the LiDAR point cloud data to obtain ground points;
generating a digital elevation model according to the ground points;
performing normalization processing on the LiDAR point cloud data by adopting the digital elevation model to obtain normalized LiDAR point cloud data;
converting the normalized LiDAR point cloud data into raster data with a preset resolution ratio to obtain tree height raster data;
acquiring multispectral data of an area to be measured, and calculating vegetation index distribution data of the area according to the multispectral data;
and estimating the forest biomass of the area to be measured by using the forest biomass, the tree height grid data and the vegetation index distribution data through a biomass regression formula.
Has the advantages that: according to the method for estimating the ground biomass by fusing the multispectral data and the LiDAR data of the unmanned aerial vehicle, the spectral information of the multispectral data of the unmanned aerial vehicle and the forest three-dimensional structure information of the LiDAR data of the unmanned aerial vehicle are integrated, inversion estimation is carried out by using real data of actual measurement points, and the inversion precision is greatly improved.
Drawings
FIG. 1 is a flow chart of a forest biomass estimation method according to a preferred embodiment of the present invention.
FIG. 2 is a functional block diagram of a preferred embodiment of the forest biomass estimation system of the present invention.
Fig. 3 is a schematic diagram of a system architecture for ground biomass estimation using multi-spectral and lidar data from drones.
Fig. 4 is a graph of the reflectance of the tree of fig. 3 over different wavelength ranges.
FIG. 5 is a schematic diagram of identification of a geographic coordinate control point of aerial photography of an area to be measured in the embodiment.
FIG. 6 is a diagram illustrating image stitching of a region under test in an embodiment.
FIG. 7 is a green band radiation calibration chart of the unmanned aerial vehicle data image of the area to be measured in the embodiment.
FIG. 8 is a near-infrared band radiation correction chart of the unmanned aerial vehicle data image of the area to be measured in the embodiment.
FIG. 9 is a red-band radiation calibration chart of the unmanned aerial vehicle data image of the area to be measured in the embodiment.
Fig. 10 is a red-band radiation correction chart of the unmanned aerial vehicle data image of the area to be measured in the embodiment.
Fig. 11 is a distribution diagram of the NDVI regions to be measured in the embodiment.
FIG. 12 is a LiDAR point cloud diagram before denoising of a region under test in an embodiment.
FIG. 13 is a LiDAR point cloud diagram after de-noising of the region under test in the embodiment.
Fig. 14 is a bottom surface point diagram after filtering of the region to be measured in the embodiment.
FIG. 15 is a diagram showing filtered non-nadirs of a region under test in an embodiment.
FIG. 16 is a 0.5m DEM map generated for the area under test in the example.
Fig. 17 is a point cloud chart before normalization of the region to be measured in the embodiment.
Fig. 18 is a point cloud image of the region to be measured after normalization in the embodiment.
Fig. 19 is a seed point display diagram generated by the layer stacking algorithm for the region to be tested in the embodiment.
Fig. 20 is a point cloud single tree segmentation result display diagram based on seed points in the region to be measured in the embodiment.
Fig. 21 is a point cloud single tree segmentation result display diagram optimized for fig. 20.
FIG. 22 is a grid diagram of tree heights of the area to be measured in the example.
Fig. 23 is a diagram showing grid registration control points of the NDVI region under test in the embodiment.
FIG. 24 is a graph showing biomass indexes of regions to be measured in examples.
Fig. 25 is a graph of the regression result calculated by the regression formula.
FIG. 26 is an estimated biomass distribution diagram of a region to be measured in the example.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention clearer and clearer, the present invention is further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
Referring to FIG. 1, the present disclosure provides some embodiments of a forest biomass estimation method.
As shown in FIG. 1, the forest biomass estimation method of the invention comprises the following steps:
s100, calculating the forest biomass of the sampling point by obtaining forest parameters of the area to be measured.
Selecting a typical sample area in an area to be measured, carrying out field biomass investigation on arbors, small arbors, grasslands, buildings and the like in the sample area, and recording the biomass in the sample area.
S200, acquiring LiDAR point cloud data of a region to be measured, and separating the LiDAR point cloud data to obtain ground points; and generating a digital elevation model according to the ground points.
Specifically, utilize unmanned aerial vehicle to carry on 1 laser radar scanner of platform, scan the region that awaits measuring, obtain the LiDAR point cloud data in this region, because laser radar scanning equipment's precision, the characteristic of testee and surrounding environmental factor all can make the appearance of noise point. Noise in LiDAR is largely divided into high-altitude noise and low-altitude noise. The high-altitude noise mainly comes from coarse particles in the air and flying objects in the air, such as birds and the like. The low noise is mainly due to multipath effects in the measurements. Because noise can affect subsequent data processing, the raw LiDAR point cloud data must be denoised before processing.
And filtering the denoised LiDAR point cloud data, and separating ground points from non-ground points from massive point cloud data. And calculating ground points by using an irregular triangulation network method to generate a Digital Elevation Model (DEM). By adopting the irregular triangulation network method, most of ground points can be effectively removed, and the advantages of high efficiency and good robustness can be achieved.
S300, carrying out normalization processing on the LiDAR point cloud data by adopting the digital elevation model to obtain normalized LiDAR point cloud data.
Specifically, in order to eliminate the influence of the terrain on the point cloud data processing process, normalization processing needs to be performed on the denoised and filtered point cloud, and the normalization of the point cloud is performed by subtracting the height of the DEM generated by the ground points from the denoised point cloud data.
S400, converting the normalized LiDAR point cloud data into raster data with a preset resolution ratio to obtain tree height raster data.
Specifically, the step S400 includes:
s410, carrying out normalization processing on the LiDAR point cloud data to obtain the absolute height of the tree to the ground;
s420, carrying out horizontal layering on the normalized LiDAR point cloud data according to preset layering setting; the preset layering setting means that layering is carried out at certain layering intervals from a specified position;
s430, carrying out point cloud clustering on each layer obtained by layering by using a K-Means clustering method to obtain a cluster;
s440, setting cluster polygons, and stacking the cluster polygons to obtain a stack diagram;
s450, identifying a local maximum value in the stacking graph by adopting a window with a fixed size to obtain a single-wood parameter;
and S460, taking the single tree parameters as seed points, performing point cloud single tree segmentation on the seed points by adopting a PCS algorithm, and editing segmentation results to obtain tree height grid data.
In the invention, normalized LiDAR point Cloud data is segmented based on layer stacking seed points, a local maximum value is identified as a seed point, then point Cloud segmentation is carried out based on the seed point by adopting a PCS (Point Cloud segmentation) algorithm, and finally each independent tree is obtained.
And (1) before segmentation, normalizing the original point cloud, eliminating the influence of the terrain on the segmentation, and obtaining the absolute height of the tree to the ground. (2) And horizontally layering the point cloud, starting from 0.5m and taking 1 m as layering interval until the point cloud is layered to the highest point. (3) Point cloud clustering, applying a clustering algorithm to each layer in order to remove some unwanted low vegetation. In addition, the bottom 3 layers need to be density scanned first, aggregating points into clusters by density and user defined minimum number of points per cluster. The CHM grid at 1 meter resolution was smoothed with a 3 x 3 pixel window and the local maxima were identified as tree heights using a 3 meter fixed window radius. The local maximum value is used as a seed point, each layer uses a K-Means clustering method, the point cloud is clustered to the seed point nearest to the point cloud, the center of the cluster is recalculated to be used as a new seed point, then all the point clouds are clustered to the seed point nearest to the point cloud, and the process is repeated iteratively until the position of the seed point is not changed any more. (4) Clustering polygons, and setting a polygon buffer area of 0.5m around each cluster. There are two main purposes, firstly, that a point other than 0.5m from the main cluster may be misclassified in that cluster and separated from another cluster. Second, as a means of connecting points and vectoring clusters, the size of the buffer can be determined qualitatively by trial and error and visual assessment to determine the optimal size of the crown, and the optimal size may be slightly different depending on the pulse density and forest type. (5) Overlapping wien maps, stacking each layer of polygons, generate a large number of rasterized overlapping polygons of 0.5 meter resolution. The overlap map defines a high density region in the crown, and multiple polygons overlap to indicate the presence of a tree. (6) The stack is smoothed, with each layer being smoothed with a 1.5 meter window size after stacking. (7) Local maxima are identified and detected in the stacked map using a fixed window size of 1.5 meters, and these local maxima represent the center of the tree, which is the point of highest overlap in the entire canopy.
Further, the segmentation method requires input of several important parameters, first, the resolution of CHM is related to pulse interval and point cloud density. The density of the data point cloud is about 10 pts/m2, so the invention adopts CHM with 0.5m resolution. The minimum tree spacing is also important and is set according to the average tree spacing under different forest conditions. In addition, the Gaussian smoothing degree can influence the divided trees, the smoothing degree is higher when the smoothing factor is larger, and the smoothing degree is smaller when the Gaussian smoothing factor is smaller. If the smoothness degree is too high, under-segmentation is easy to occur, and if the smoothness degree is too low, over-segmentation is easy to occur, so that the size of the Gaussian smoothing factor is important. The smoothing radius, i.e. the size of the window used for gaussian smoothing, should be comparable to the average crown diameter. After segmentation, parameters such as the position of a single tree, the height of the tree, the diameter of a crown, the area of the crown, the volume of the crown and the like can be obtained. And converting the point cloud after drying removal and ground point separation and normalization into raster data with the resolution of 0.5m, namely the tree height raster data.
S500, acquiring multispectral data of an area to be measured, and calculating vegetation index distribution data of the area according to the multispectral data;
in particular, since LiDAR data provides only three-dimensional structural information of the ground, vegetation growth information having a strong correlation with biomass is difficult to provide. Therefore, the method and the device describe the growth condition information of the vegetation by combining the normalized vegetation index calculated by the multispectral data of the unmanned aerial vehicle.
The normalized vegetation index (NDVI) is calculated by the formula:
wherein,andrespectively representing the reflectivity of the near infrared band and the red light band.
The normalized vegetation index (NDVI) is one of important parameters reflecting the growth and nutrition information of crops, has strong correlation with biomass, and is often used for large-area ground biomass estimation.
S600, estimating the forest biomass of the area to be measured by using the forest biomass, the tree height grid data and the vegetation index distribution data through a biomass regression formula.
The biomass regression formula is:
wherein AGB is ground biomass, H is tree height, and NDVI is a normalized vegetation index. From this, a new index, the biomass index (AGBI), is derived, which has a higher correlation with the ground biomass, as shown in the following formula:
therefore, the formula for fitting the ground biomass to the biomass index is as follows:
the invention also provides a preferred embodiment of the forest biomass estimation system, which comprises the following steps:
as shown in fig. 2, the forest biomass estimation system according to the embodiment of the present invention includes: a processor 10, and a memory 20 connected to said processor 10,
the memory 20 stores a forest biomass estimation program which, when executed by the processor 10, performs the steps of:
the memory stores a terrestrial biomass estimation program that when executed by the processor performs the steps of:
acquiring forest biomass of sampling points in an area to be detected;
the method comprises the steps of obtaining LiDAR point cloud data of a region to be measured, and separating the LiDAR point cloud data to obtain ground points;
generating a digital elevation model according to the ground points;
performing normalization processing on the LiDAR point cloud data by adopting the digital elevation model to obtain normalized LiDAR point cloud data;
converting the normalized LiDAR point cloud data into raster data with a preset resolution ratio to obtain tree height raster data;
acquiring multispectral data of an area to be measured, and calculating vegetation index distribution data of the area according to the multispectral data;
and estimating the forest biomass of the area to be measured by using the forest biomass, the tree height grid data and the vegetation index distribution data through a biomass regression formula.
The forest biomass estimation method and system provided by the invention are further explained by an embodiment as follows:
the system architecture of the embodiment comprises 1 unmanned aerial vehicle carrying platform, 1 laser radar scanner, 1 multispectral image sensor and 1 GPS receiver, and is shown in fig. 3. LiDAR point cloud data and multispectral image data are obtained through a multispectral image sensor of an unmanned airborne laser radar scanner. When the multispectral image sensor acquires multispectral image data, the vegetation reflects light of different wave bands, and the reflectivity is shown in fig. 4.
1. The unmanned plane multispectral image data processing flow is as follows:
(1) unmanned aerial vehicle data image concatenation geometric correction, as shown in figure 5 according to the geographical coordinate control point of taking photo by plane, splices unmanned aerial vehicle data image, and the image result is shown in figure 6.
(2) The unmanned aerial vehicle data image radiation correction comprises green band radiation correction, near infrared band correction, red band correction and red band correction, and the correction results are shown in figures 7-10.
(3) The normalized vegetation index (NDVI) was calculated from the drone image data, with the results shown in fig. 11.
2. The processing flow of the laser point cloud data of the unmanned aerial vehicle is as follows:
(1) air noise points are removed, and data quality is improved; as shown in FIG. 12, which is a pre-denoised LiDAR point cloud image, FIG. 13 is a post-denoised LiDAR point cloud image.
(2) Ground points are separated from the point cloud data, which is a filtered ground point map as shown in fig. 14, and a filtered non-ground point map as shown in fig. 15.
(3) DEM is generated based on ground points, and FIG. 16 is a generated 0.5m DEM diagram
(4) And normalizing the point cloud by using the DEM to remove the influence of the terrain, as shown in FIG. 17, the point cloud data after being denoised is not subjected to normalization processing. FIG. 18 is the point cloud data after normalization processing for the left image, where the Z value of each point is the true vertical distance from the ground, and when the point is at the top of the tree, its Z value is the tree height.
(5) And (3) generating seed points by layer stacking, extracting the single-tree positions (the seed point file is a comma-separated CSV file which comprises four columns, namely tree ID, X, Y and Z coordinates in sequence) from the normalized point cloud data by adopting a layer stacking algorithm, and carrying out single-tree segmentation on the point cloud by taking the information as the seed points. As shown in fig. 19.
(6) And performing single-tree segmentation on the point cloud based on the seed points to obtain the position of the single tree, the height of the tree, the diameter of the crown, the area of the crown and the volume of the crown. The single wood segmentation results are shown in fig. 20.
(7) The results of the single tree segmentation are checked by the ALS editing tool, meanwhile, manual interactive editing such as addition and deletion is performed on the seed points, and the point cloud is segmented again based on the edited seed points, and the results are shown in fig. 21.
(8) And converting the ground points subjected to drying removal and separation and the point cloud subjected to normalization into raster data with the resolution of 0.5m, namely obtaining a tree height raster image. As shown in fig. 22.
(9) And accurately registering the NDVI grid map by using the tree height grid map. As shown in fig. 23.
(10) According to X, Y coordinates of the biomass survey point, a corresponding seed point file is configured in csv format, as shown in the following table one.
Watch 1
| TreeID | TreeLocationX | TreeLocationY | TreeLocationZ |
| 1 | 225397.939 | 2521767.749 | 5 |
| 2 | 225404.607 | 2521766.955 | 6 |
| 3 | 225420.72 | 2521735.919 | 7 |
| 4 | 225486.767 | 2521680.314 | 8 |
| 5 | 225618.834 | 2521571.57 | 10 |
| 6 | 225759.509 | 2521099.765 | 12 |
| 7 | 225687.277 | 2521068.015 | 13 |
| 8 | 225284.078 | 2521420.229 | 6 |
(11) The seed point file is opened in the lidar360, and the nearest single tree height (or the most appropriate tree height, which needs manual inspection) adjacent to the seed points is read according to the seed points and the point cloud after the single tree segmentation (such as the position of the single tree, the height of the single tree, and the like can be obtained). As shown in table two.
Watch two
| TreeID | TreeLocationX | TreeLocationY | NDVI | Bio | TreeHeight |
| 1 | 225397.939 | 2521767.749 | 0.185 | 1 | 0.377 |
| 2 | 225404.607 | 2521766.955 | -0.207 | 0 | 0 |
| 3 | 225420.72 | 2521735.919 | 0.729 | 600 | 7.204 |
| 4 | 225486.767 | 2521680.314 | 0.548 | 30 | 4.297 |
| 5 | 225618.834 | 2521571.57 | 0.762 | 300 | 6.283 |
| 6 | 225759.509 | 2521099.765 | 0.788 | 1000 | 6.122 |
| 7 | 225687.277 | 2521068.015 | 0.689 | 200 | 4.758 |
| 8 | 225284.078 | 2521420.229 | 0.467 | 10 | 0.061 |
2. Biomass calculation flow
(1) And multiplying the tree height grid map and the registered NDVI map to obtain a biomass index map. As shown in fig. 24.
(2) In-situ biomass sampling
More than 10 samples were collected in the field in the study area, and the samples of field biomass were as follows:
weeds; geographic coordinates: 22.782071,114.334434, respectively; biomass: 0.50 kg; eucalyptus, 12 meters in height and 30 centimeters in diameter at breast height; geographic coordinates: 22.774872,114.328818, respectively; biomass: about 680 kg; small arbor, height 2 m, diameter at breast height 8 cm; geographic coordinates: 22.783026,114.326606, respectively; biomass: about 30 kg.
(3) And (4) performing regression according to the on-site biomass investigation value and the biomass index value, and calculating a regression formula. The regression results are shown in FIG. 25.
(4) The biomass was estimated according to the regression formula and plotted as shown in FIG. 26.
In summary, the method for estimating forest biomass and the system thereof provided by the invention comprise the following steps: calculating the forest biomass of the sampling point by acquiring forest parameters of the area to be measured; the method comprises the steps of obtaining LiDAR point cloud data of a region to be measured, and separating the LiDAR point cloud data to obtain ground points; generating a digital elevation model according to the ground points; performing normalization processing on the LiDAR point cloud data by adopting the digital elevation model to obtain normalized LiDAR point cloud data; converting the normalized LiDAR point cloud data into raster data with a preset resolution ratio to obtain tree height raster data; acquiring multispectral data of an area to be measured, and calculating vegetation index distribution data of the area according to the multispectral data; and estimating the forest biomass of the area to be measured by using the forest biomass, the tree height grid data and the vegetation index distribution data through a biomass regression formula. According to the method for estimating the ground biomass by fusing the multispectral data and the LiDAR data of the unmanned aerial vehicle, the spectral information of the multispectral data of the unmanned aerial vehicle and the forest three-dimensional structure information of the LiDAR data of the unmanned aerial vehicle are integrated, inversion estimation is carried out by using real data of actual measurement points, and the inversion precision is greatly improved.
It is to be understood that the invention is not limited to the examples described above, but that modifications and variations may be effected thereto by those of ordinary skill in the art in light of the foregoing description, and that all such modifications and variations are intended to be within the scope of the invention as defined by the appended claims.
Claims (10)
1. A forest biomass estimation method is characterized by comprising the following steps:
calculating the forest biomass of the sampling point by acquiring forest parameters of the area to be measured;
the method comprises the steps of obtaining LiDAR point cloud data of a region to be measured, and separating the LiDAR point cloud data to obtain ground points; generating a digital elevation model according to the ground points;
performing normalization processing on the LiDAR point cloud data by adopting the digital elevation model to obtain normalized LiDAR point cloud data;
converting the normalized LiDAR point cloud data into raster data with a preset resolution ratio to obtain tree height raster data;
acquiring multispectral data of an area to be measured, and calculating vegetation index distribution data of the area according to the multispectral data;
and estimating the forest biomass of the area to be measured by using the forest biomass, the tree height grid data and the vegetation index distribution data through a biomass regression formula.
2. The forest biomass estimation method according to claim 1, wherein the step of obtaining multispectral data of a region to be measured and calculating vegetation index distribution data of the region according to the multispectral data specifically comprises:
according to the aerial photography geographic coordinate control points, carrying out image splicing on unmanned aerial vehicle data images of the area to be tested;
and carrying out radiation correction on the spliced unmanned aerial vehicle data image, and calculating a normalized vegetation index of vegetation in the region to be detected.
3. The forest biomass estimation method according to claim 1, wherein the step of converting the normalized LiDAR point cloud data into raster data of a predetermined resolution to obtain tree height raster data includes:
carrying out normalization processing on the LiDAR point cloud data to obtain the absolute height of the tree to the ground;
carrying out horizontal layering on the normalized LiDAR point cloud data according to a preset layering setting; the preset layering setting means that layering is carried out at certain layering intervals from a specified position;
carrying out point cloud clustering on each layer obtained by layering by using a K-Means clustering method to obtain a cluster;
setting cluster polygons, and stacking the cluster polygons to obtain a stack diagram;
identifying a local maximum value in the stacking graph by adopting a window with a fixed size to obtain a single tree parameter;
and taking the single tree parameters as seed points, performing point cloud single tree segmentation on the seed points by adopting a PCS algorithm, and editing the segmentation result to obtain tree height grid data.
4. The forest biomass estimation method of claim 1, wherein the biomass regression formula is:
wherein AGB is ground biomass, H is tree height, and NDVI is a normalized vegetation index.
5. The forest biomass estimation method according to claim 2, wherein the step of performing radiation correction on the spliced unmanned aerial vehicle data image to calculate a normalized vegetation index of vegetation in the area to be measured, wherein the radiation correction includes: green band correction, near infrared band correction, red edge band correction, and red band correction.
6. The forest biomass estimation method according to claim 2, wherein the normalized vegetation index is calculated by the formula:
wherein,andrespectively representing the reflectivity of the near infrared band and the red light band.
7. The forest biomass estimation method according to claim 1, wherein the predetermined resolution is a resolution of 0.3-0.6 meters.
8. The forest biomass estimation method according to claim 3, wherein the designated position is 0.5-0.7 meters, and the layering interval is 1 meter.
9. The forest biomass estimation method according to claim 1, wherein the fixed size is 1.0-2.0 meters.
10. A surface biomass estimation system, comprising: a processor, and a memory coupled to the processor;
the memory stores a terrestrial biomass estimation program that when executed by the processor performs the steps of:
acquiring forest biomass of sampling points in an area to be detected;
the method comprises the steps of obtaining LiDAR point cloud data of a region to be measured, and separating the LiDAR point cloud data to obtain ground points;
generating a digital elevation model according to the ground points;
performing normalization processing on the LiDAR point cloud data by adopting the digital elevation model to obtain normalized LiDAR point cloud data;
converting the normalized LiDAR point cloud data into raster data with a preset resolution ratio to obtain tree height raster data;
acquiring multispectral data of an area to be measured, and calculating vegetation index distribution data of the area according to the multispectral data;
and estimating the forest biomass of the area to be measured by using the forest biomass, the tree height grid data and the vegetation index distribution data through a biomass regression formula.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910295160.6A CN110148116A (en) | 2019-04-12 | 2019-04-12 | A kind of forest biomass evaluation method and its system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910295160.6A CN110148116A (en) | 2019-04-12 | 2019-04-12 | A kind of forest biomass evaluation method and its system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110148116A true CN110148116A (en) | 2019-08-20 |
Family
ID=67588570
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910295160.6A Pending CN110148116A (en) | 2019-04-12 | 2019-04-12 | A kind of forest biomass evaluation method and its system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110148116A (en) |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110781862A (en) * | 2019-11-07 | 2020-02-11 | 航天信德智图（北京）科技有限公司 | Multispectral image-based forest arbor and shrub identification method |
| CN110823813A (en) * | 2019-11-19 | 2020-02-21 | 深圳市智绘科技有限公司 | Forest land ground biomass estimation method |
| CN111462134A (en) * | 2020-03-31 | 2020-07-28 | 武汉大学 | Single-tree segmentation method and system for fusing high-resolution remote sensing image and laser radar point cloud |
| CN111489388A (en) * | 2020-04-20 | 2020-08-04 | 黑龙江工程学院 | Biomass inversion method based on effective crown information |
| CN111754515A (en) * | 2019-12-17 | 2020-10-09 | 北京京东尚科信息技术有限公司 | Method and device for sequential gripping of stacked articles |
| CN112215169A (en) * | 2020-10-10 | 2021-01-12 | 华中农业大学 | Crop plant height and biomass self-adaptive high-precision resolving method based on low-altitude unmanned-machine passive remote sensing |
| CN112434617A (en) * | 2020-11-26 | 2021-03-02 | 南京观微空间科技有限公司 | Forest biomass change monitoring method and system based on multi-source remote sensing data |
| CN112560661A (en) * | 2020-12-10 | 2021-03-26 | 首都师范大学 | Leaf area index calculation method and device, electronic equipment and readable storage medium |
| CN112730306A (en) * | 2020-12-21 | 2021-04-30 | 合肥工业大学智能制造技术研究院 | Cloud cluster three-dimensional space distribution measuring method, device and system |
| CN114120158A (en) * | 2021-11-29 | 2022-03-01 | 四川省自然资源科学研究院 | Leaf area statistical method and device based on ultra-low altitude remote sensing and storage medium |
| CN114463512A (en) * | 2021-12-24 | 2022-05-10 | 广州极飞科技股份有限公司 | Point cloud data processing method, vectorization method and device |
| CN115629393A (en) * | 2022-11-30 | 2023-01-20 | 北京市农林科学院智能装备技术研究中心 | Biomass estimation method, device, system, electronic device and storage medium |
| CN116127327A (en) * | 2023-04-07 | 2023-05-16 | 广东省科学院广州地理研究所 | Forest ground biomass inversion method, device, equipment and storage medium |
| CN117368118A (en) * | 2023-10-09 | 2024-01-09 | 太原理工大学 | Mining area biomass monitoring method based on multispectral and point cloud data processing |
| JP7576721B2 (en) | 2023-03-06 | 2024-10-31 | 四川省林業和草原調査規劃院（四川省林業和草原生態環境監測中心） | Dynamic accounting method for carbon sinks of afforestation and forest management carbon sink projects |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102393180A (en) * | 2011-10-19 | 2012-03-28 | 中国林业科学研究院资源信息研究所 | Method for automatically extracting forest stand upper layer tree parameters from LiDAR point cloud data |
| CN104778451A (en) * | 2015-03-31 | 2015-07-15 | 中国科学院上海技术物理研究所 | Grassland biomass remote sensing inversion method considering grassland height factor |
| CN107085710A (en) * | 2017-04-26 | 2017-08-22 | 长江空间信息技术工程有限公司(武汉) | A kind of single wooden extraction method based on multispectral LiDAR data |
| CN108020211A (en) * | 2017-12-01 | 2018-05-11 | 云南大学 | A kind of method of unmanned plane aeroplane photography estimation instruction plant biomass |
| US20190034725A1 (en) * | 2016-01-29 | 2019-01-31 | Global Surface Intelligence Limited | System and method for earth observation and analysis |
-
2019
- 2019-04-12 CN CN201910295160.6A patent/CN110148116A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102393180A (en) * | 2011-10-19 | 2012-03-28 | 中国林业科学研究院资源信息研究所 | Method for automatically extracting forest stand upper layer tree parameters from LiDAR point cloud data |
| CN104778451A (en) * | 2015-03-31 | 2015-07-15 | 中国科学院上海技术物理研究所 | Grassland biomass remote sensing inversion method considering grassland height factor |
| US20190034725A1 (en) * | 2016-01-29 | 2019-01-31 | Global Surface Intelligence Limited | System and method for earth observation and analysis |
| CN107085710A (en) * | 2017-04-26 | 2017-08-22 | 长江空间信息技术工程有限公司(武汉) | A kind of single wooden extraction method based on multispectral LiDAR data |
| CN108020211A (en) * | 2017-12-01 | 2018-05-11 | 云南大学 | A kind of method of unmanned plane aeroplane photography estimation instruction plant biomass |
Non-Patent Citations (1)
| Title |
|---|
| 杨秋丽: "机载LiDAR单木分割方法对比及其影响因素分析", 《中国优秀硕士学位论文全文数据库 农业科技辑》 * |
Cited By (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110781862A (en) * | 2019-11-07 | 2020-02-11 | 航天信德智图（北京）科技有限公司 | Multispectral image-based forest arbor and shrub identification method |
| CN110823813A (en) * | 2019-11-19 | 2020-02-21 | 深圳市智绘科技有限公司 | Forest land ground biomass estimation method |
| CN111754515B (en) * | 2019-12-17 | 2024-03-01 | 北京京东乾石科技有限公司 | Sequential gripping method and device for stacked articles |
| CN111754515A (en) * | 2019-12-17 | 2020-10-09 | 北京京东尚科信息技术有限公司 | Method and device for sequential gripping of stacked articles |
| CN111462134A (en) * | 2020-03-31 | 2020-07-28 | 武汉大学 | Single-tree segmentation method and system for fusing high-resolution remote sensing image and laser radar point cloud |
| CN111489388A (en) * | 2020-04-20 | 2020-08-04 | 黑龙江工程学院 | Biomass inversion method based on effective crown information |
| CN112215169A (en) * | 2020-10-10 | 2021-01-12 | 华中农业大学 | Crop plant height and biomass self-adaptive high-precision resolving method based on low-altitude unmanned-machine passive remote sensing |
| CN112434617A (en) * | 2020-11-26 | 2021-03-02 | 南京观微空间科技有限公司 | Forest biomass change monitoring method and system based on multi-source remote sensing data |
| CN112434617B (en) * | 2020-11-26 | 2021-08-13 | 南京观微空间科技有限公司 | Forest biomass change monitoring method and system based on multi-source remote sensing data |
| CN112560661A (en) * | 2020-12-10 | 2021-03-26 | 首都师范大学 | Leaf area index calculation method and device, electronic equipment and readable storage medium |
| CN112730306A (en) * | 2020-12-21 | 2021-04-30 | 合肥工业大学智能制造技术研究院 | Cloud cluster three-dimensional space distribution measuring method, device and system |
| CN114120158A (en) * | 2021-11-29 | 2022-03-01 | 四川省自然资源科学研究院 | Leaf area statistical method and device based on ultra-low altitude remote sensing and storage medium |
| CN114463512A (en) * | 2021-12-24 | 2022-05-10 | 广州极飞科技股份有限公司 | Point cloud data processing method, vectorization method and device |
| CN115629393A (en) * | 2022-11-30 | 2023-01-20 | 北京市农林科学院智能装备技术研究中心 | Biomass estimation method, device, system, electronic device and storage medium |
| JP7576721B2 (en) | 2023-03-06 | 2024-10-31 | 四川省林業和草原調査規劃院（四川省林業和草原生態環境監測中心） | Dynamic accounting method for carbon sinks of afforestation and forest management carbon sink projects |
| CN116127327A (en) * | 2023-04-07 | 2023-05-16 | 广东省科学院广州地理研究所 | Forest ground biomass inversion method, device, equipment and storage medium |
| CN116127327B (en) * | 2023-04-07 | 2023-07-11 | 广东省科学院广州地理研究所 | Forest ground biomass inversion method, device, equipment and storage medium |
| CN117368118A (en) * | 2023-10-09 | 2024-01-09 | 太原理工大学 | Mining area biomass monitoring method based on multispectral and point cloud data processing |
| CN117368118B (en) * | 2023-10-09 | 2024-10-18 | 太原理工大学 | Mining area biomass monitoring method based on multispectral and point cloud data processing |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110148116A (en) | A kind of forest biomass evaluation method and its system | |
| US11615606B2 (en) | Tree crown extraction method based on unmanned aerial vehicle multi-source remote sensing | |
| Zhang et al. | Assessment of defoliation during the Dendrolimus tabulaeformis Tsai et Liu disaster outbreak using UAV-based hyperspectral images | |
| CN109146889B (en) | Farmland boundary extraction method based on high-resolution remote sensing image | |
| Simonse et al. | Automatic determination of forest inventory parameters using terrestrial laser scanning | |
| Kattenborn et al. | Automatic single tree detection in plantations using UAV-based photogrammetric point clouds | |
| US8208689B2 (en) | Method for determination of stand attributes and a computer program for performing the method | |
| CN109031344B (en) | Method for jointly inverting forest structure parameters by full-waveform laser radar and hyperspectral data | |
| CN111709981A (en) | Registration method of laser point cloud and analog image with characteristic line fusion | |
| Zhang et al. | Extraction of tree crowns damaged by Dendrolimus tabulaeformis Tsai et Liu via spectral-spatial classification using UAV-based hyperspectral images | |
| CN112669363B (en) | Method for measuring three-dimensional green space of urban green space | |
| CN112396019A (en) | Vegetation distribution identification method and system based on unmanned aerial vehicle and readable storage medium | |
| Aubry-Kientz et al. | Multisensor data fusion for improved segmentation of individual tree crowns in dense tropical forests | |
| Bandyopadhyay et al. | Classification and extraction of trees and buildings from urban scenes using discrete return LiDAR and aerial color imagery | |
| US10810704B2 (en) | Method for processing an optical image acquired in the presence of aerosols and/or clouds in the atmosphere | |
| Smits et al. | Individual tree identification using different LIDAR and optical imagery data processing methods | |
| CN115453555A (en) | Unmanned aerial vehicle rapid monitoring method and system for grassland productivity | |
| Wierzbicki et al. | Method of radiometric quality assessment of NIR images acquired with a custom sensor mounted on an unmanned aerial vehicle | |
| Grishin et al. | Tree inventory with LiDAR data | |
| Alba et al. | Filtering vegetation from terrestrial point clouds with low-cost near infrared cameras | |
| Yin et al. | Individual tree parameters estimation for chinese fir (cunninghamia lanceolate (lamb.) hook) plantations of south china using UAV Oblique Photography: Possibilities and Challenges | |
| CN118212556A (en) | Urban ecological corridor internal constituent element screening method combining active and passive remote sensing | |
| CN118072160A (en) | Single wood scale biomass accurate estimation method and system based on improved Mask R-CNN | |
| Hui et al. | Individual tree extraction from UAV lidar point clouds based on self-adaptive Mean shift segmentation | |
| CN111563472A (en) | Method and device for rapidly extracting tobacco plume forest land burned area |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190820 |