CN110598707A - Single tree crown segmentation method for water pouring spread and energy function control of airborne laser point cloud - Google Patents
Single tree crown segmentation method for water pouring spread and energy function control of airborne laser point cloud Download PDFInfo
- Publication number
- CN110598707A CN110598707A CN201910632238.9A CN201910632238A CN110598707A CN 110598707 A CN110598707 A CN 110598707A CN 201910632238 A CN201910632238 A CN 201910632238A CN 110598707 A CN110598707 A CN 110598707A
- Authority
- CN
- China
- Prior art keywords
- crown
- water
- tree
- height
- spread
- 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.)
- Granted
Links
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 title claims abstract description 153
- 238000000034 method Methods 0.000 title claims abstract description 59
- 230000011218 segmentation Effects 0.000 title claims abstract description 26
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 49
- 238000003892 spreading Methods 0.000 claims abstract description 43
- 239000013598 vector Substances 0.000 claims description 17
- 230000007423 decrease Effects 0.000 claims description 6
- 238000012804 iterative process Methods 0.000 claims description 4
- 238000013461 design Methods 0.000 claims description 3
- 238000001514 detection method Methods 0.000 abstract description 25
- 238000004590 computer program Methods 0.000 abstract description 5
- 241000218645 Cedrus Species 0.000 abstract description 4
- 244000166124 Eucalyptus globulus Species 0.000 abstract 3
- 241000219927 Eucalyptus Species 0.000 description 53
- 210000004027 cell Anatomy 0.000 description 26
- 241000894007 species Species 0.000 description 14
- 238000000926 separation method Methods 0.000 description 11
- 241000196324 Embryophyta Species 0.000 description 10
- 230000000694 effects Effects 0.000 description 10
- 238000010586 diagram Methods 0.000 description 8
- 238000011160 research Methods 0.000 description 8
- 239000002023 wood Substances 0.000 description 8
- 238000009826 distribution Methods 0.000 description 6
- 238000009499 grossing Methods 0.000 description 6
- 230000001360 synchronised effect Effects 0.000 description 6
- 238000004364 calculation method Methods 0.000 description 5
- 244000050510 Cunninghamia lanceolata Species 0.000 description 4
- 238000005259 measurement Methods 0.000 description 4
- 210000003888 boundary cell Anatomy 0.000 description 3
- 238000011156 evaluation Methods 0.000 description 3
- 238000011065 in-situ storage Methods 0.000 description 3
- 238000012360 testing method Methods 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 238000005520 cutting process Methods 0.000 description 2
- 238000011835 investigation Methods 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 239000000203 mixture Substances 0.000 description 2
- 230000000877 morphologic effect Effects 0.000 description 2
- 239000002689 soil Substances 0.000 description 2
- DSSYKIVIOFKYAU-XCBNKYQSSA-N (R)-camphor Chemical compound C1C[C@@]2(C)C(=O)C[C@@H]1C2(C)C DSSYKIVIOFKYAU-XCBNKYQSSA-N 0.000 description 1
- 241001270131 Agaricus moelleri Species 0.000 description 1
- CIWBSHSKHKDKBQ-JLAZNSOCSA-N Ascorbic acid Chemical compound OC[C@H](O)[C@H]1OC(=O)C(O)=C1O CIWBSHSKHKDKBQ-JLAZNSOCSA-N 0.000 description 1
- 241000723346 Cinnamomum camphora Species 0.000 description 1
- 244000103090 Eucalyptus robusta Species 0.000 description 1
- 240000007472 Leucaena leucocephala Species 0.000 description 1
- 235000010643 Leucaena leucocephala Nutrition 0.000 description 1
- 241000218378 Magnolia Species 0.000 description 1
- 241001106462 Ulmus Species 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 229960000846 camphor Drugs 0.000 description 1
- 229930008380 camphor Natural products 0.000 description 1
- 238000012512 characterization method Methods 0.000 description 1
- 230000000052 comparative effect Effects 0.000 description 1
- 230000002860 competitive effect Effects 0.000 description 1
- 230000001276 controlling effect Effects 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000035558 fertility Effects 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000002955 isolation Methods 0.000 description 1
- 238000002156 mixing Methods 0.000 description 1
- 235000015816 nutrient absorption Nutrition 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000001556 precipitation Methods 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 230000000630 rising effect Effects 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 238000000638 solvent extraction Methods 0.000 description 1
- 238000005507 spraying Methods 0.000 description 1
- 235000005221 swamp mahogany Nutrition 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
- G06V10/267—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion by performing operations on regions, e.g. growing, shrinking or watersheds
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Length Measuring Devices By Optical Means (AREA)
- Image Processing (AREA)
Abstract
The patent provides a crown segmentation algorithm based on airborne laser radar data, the boundary of a crown area is described by a water spreading method on a Digital Surface Model (DSM), an energy function for controlling water spreading is established by combining the gradient direction on the DSM and the height difference between a crown boundary unit and the top points of the two nearest crowns, the limitation of synchronously realizing water spreading in each crown by a computer program is made up, and the segmentation result of an overlapping area between adjacent crowns is optimized. The results of the experiment show that the average call rate (recal) of the cedar in the mixed forest is 0.90, the accuracy (precision) of the call rate is 0.71, the detected overall accuracy (f) of the tree is 0.80, and for the eucalyptus, the call is 94.76%, the precision is 0.81, and the f is 0.86. In all plots, the average detection precision of the fir crown diameter is lower than that of eucalyptus, and the detection precision of the fir tree height is 0.48% higher than that of eucalyptus.
Description
Technical Field
The invention relates to a method for extracting a single tree profile from airborne laser radar data.
Background
A great deal of tree data can be obtained by scanning the forest trees by adopting the airborne laser radar. The singular data are data of the whole forest, and are contour data of individual trees which are often needed when mechanical work such as accurate target spraying is carried out. How to process the laser point cloud data acquired by the airborne laser radar and obtain the outline data of a single tree is difficult to solve.
Disclosure of Invention
The invention aims to provide a single tree crown segmentation method for water pouring spread and energy function control of airborne laser point cloud, which can process laser radar data to obtain real and effective single tree contour data.
The invention relates to a single tree crown segmentation method for water pouring spread and energy function control of airborne laser point cloud, which comprises the following steps:
a. converting tree point cloud data acquired by an airborne laser radar into a digital surface model C; rasterizing the digital surface model C of the forest stand to convert the digital surface model C into a plurality of cells Ci,jA combination of (1); then, a local maximum algorithm is utilized to search a unit C corresponding to the local maximum of the z value from the digital surface model Ci,jFurther detecting the crown top point;
b. taking the crown top as a seed point of a water pouring spreading algorithm, pouring water into the whole crown forest stand, and using the water in each crown to spread synchronously to draw the boundary of the tree;
in the method for segmenting the single plant crown, the specific process of the step b is as follows:
inverting the digital surface model C of the tree, regarding each spatial crown as a pit, and layering the inverted digital surface model C according to the height interval;
pouring water, wherein the water naturally flows to the lowest layer of the pit;
continuing pouring water to cause the water level to rise and the area to grow synchronously, starting spreading of the water, dividing the edges of all the crowns, and taking edge information corresponding to lower-layer information filled with the water as guidance to accurately separate the upper intersection areas of the adjacent crowns;
and (5) finishing the water pouring process and describing the spatial outline of each tree.
In the method for cutting the single plant crown, the water spreading process comprises the following steps:
after the digital surface model C is divided into different levels by height, region growing starts at the highest level l detected1Crown vertex t of treekThese crown vertices tkCell C located on digital surface model Ci,jIn (c) thati,j＝＝tk,k∈[1,2,...r]&tk∈l1(ii) a Then combining the phenotypic characteristics of the shape of the crown, namely, the z value of the scanning laser point on the surface of the crown is gradually reduced from the same polarity at the top of the crown to each direction, and simulating water spreading by adopting a neighborhood search strategy; when the crown vertex t is detectedkAnd its location on DSM, to find the current layer l1Each of the interior and the unit c as the crown topi,jAdjacent cell ofAnd satisfies the phenotypic characteristics of crown shapeTo help determine the course of water flow;represents each unit ci,jThe corresponding z value; the water region of each crown then continues to spread according to the crown-shaped phenotypic characteristics, fromStart and synchronously creep to wherein To representAnd satisfiesNext, the process of the present invention is described,repeating the iterative process until all the units c corresponding to the crown vertexesi,jUnits with a partnership and belonging to the current level l1All the cells are spread by water;
at a height interval of l1After the spread of the water inside is completed, the corresponding area of the digital surface model C where each pit is immersed in the water will be inherited to the next height interval l2Performing the following steps; then, at a height interval of l2The crown with the newly added crown vertex is added into the area growth, the water level growth is not completed, and the crown vertex exists in the upper layer l1Will continue their area growth and the overlapping area between the two crowns will be correctly segmented according to the water spread inherited from the upper layers; the water spreading pattern in each pit will be applied in height order to the next height interval until all height layers have been treated.
In the single tree crown segmentation method, the gradient direction information of the digital surface model C and the height difference information of each unit in the digital surface model C are combined to control the water spreading sequence, namely:
design energy function Boundarryenergy (t)k) To evaluate the correctness of the water spreading sequence in each crown, Boundaryenergy (t) during the water spreading processk) The value should continue to decrease until a minimum value is reached;
wherein ,
representing the calculated gradient vector for each cell in the digital surface model C,
indicating that the vector to the north is measured in a clockwise mannerThe included angle of (A);
q denotes the current border element c belonging to each crown in each step of the water spreadi,jThe number of (2);
representing the vertex t from the crownkA vector pointing north;
representing vectorsTo vectorClockwise angle of (d);
representing the vertex t from the crownkTo the current cell ci,jA vector of (a);means that the water area boundary unit of each crown should be away from the height difference of the two crown vertexes closest to the water area boundary unit of each crown along with the continuous water pouring processThe reciprocal of the height difference should be decreased continuously while increasing continuously, and β is the corresponding weight coefficient.
Therefore, we construct the energy function Boundarryenergy (t)k) Should be constantly expanded during pouring and water level within each crownIn the process, the calculated value is continuously reduced, namely, the energy function is reduced to the minimum value, and the accurate individual plant separation result is obtained. Meanwhile, if the energy function value suddenly increases in the process of pouring water in a single crown (the original descending trend is broken), the water level represents that the error spread occurs in a certain crown and extends to other adjacent crowns. For such a case, the water level spreading order of the wrong crown (corresponding to the sudden increase of the energy function) can be arranged to the end by adjusting the order of the water level spreading in each crown, thereby allowing the overlapping crown areas to be correctly classified into each tree body.
The invention has the beneficial effects that: the patent provides a crown segmentation algorithm based on airborne laser radar data, and a water spreading method on a Digital Surface Model (DSM) is utilized to depict the boundary of a crown area; and establishing an energy function for controlling water spread by combining the gradient direction on the DSM and the height difference between the boundary unit of the crown and the top points of the two nearest crowns, making up for the limitation of synchronously realizing the water spread in each crown by a computer program, and optimizing the segmentation result of the overlapping area between the adjacent crowns. The patent uses 12 plots of two species (fir and eucalyptus) with different crown sizes and tree heights to evaluate the efficiency of the method and compares the tree parameters obtained by the method with measured values. The results showed that the average call rate (recill) of the cedar in the mixed forest was 0.90, the accuracy of the call rate (precision) was 0.71, the detected overall accuracy (f) of the tree was 0.80, and the cedar was lower than the indexes of eucalyptus (recill 94.76%, precision 0.81, f 0.86). Average detection precision (R) of fir crown diameter in all plots2The average detection precision (R) of eucalyptus is lower than that of 0.79, RMSE is 0.25m, rRMSE is 13.56 percent20.77, RMSE 0.24m, RMSE 12.21%). Detection precision (R) of fir tree height2The detection precision of the eucalyptus is higher than 0.48 percent (R is 0.87, RMSE is 0.89m, rRMSE is 5.82 percent) of the detection precision of the eucalyptus20.91, RMSE 1.01m, RMSE 6.30%). The method has the advantages of relatively simple calculation process, high material rate and good accuracy.
Drawings
FIG. 1 is a workflow diagram of a single tree profile extracted from airborne lidar data;
FIG. 2 is a schematic diagram of the process of water spread expansion in an inverted tree model;
FIG. 3 is a program run diagram of a growth process of water spread in corresponding DSMs and crowns;
FIG. 4 is a schematic diagram of control criteria for determining a crown boundary design;
figure 5 is a trend plot of the spread sequence and the energy function value of the formation on each crown.
FIG. 6 is a graph showing the results of analysis of individual tree crowns for six selected plots.
FIG. 7 is a graph comparing crown radius values measured in situ with crown radius values obtained according to the method of the present patent.
FIG. 8 is a graph comparing the measured in situ height to the height obtained by the method of the present patent.
FIG. 9 is a graph of the effect of height separation on crown radius rRMSE and computation time.
Detailed Description
1 area of investigation
The research area is located in a peak forest, a subtropical artificial forest in Guangxi province in the south China, the range of the subtropical artificial forest is from 108 degrees 7 ' of east longitude to 10838 degrees of east longitude, and from 22 degrees 49 ' of north latitude to 23 degrees 5' of north latitude, the research area is located on the south side of the return line of north China, the subtropical artificial forest belongs to subtropical monsoon humid climate, the sunshine is sufficient, the rainfall is sufficient, the frost is less, snow does not fall, the climate is mild, the annual average air temperature is about 21.6 degrees, the precipitation is 1304 millimeters, and the altitude is 78-468 meters. The area of investigation occupied 52 square kilometers. The terrain is predominantly gentle, with a slope of about 20 to 47 m, with an average of 33. The main tree species in the research area are Eucalyptus (Eucalyptus robusta) and fir (Cunninghamialanolta).
2. Materials and methods
2.1 scheme
The workflow illustrates a technique for extracting individual tree profiles from airborne lidar data (fig. 1). Rasterizing a Digital Surface Model (DSM) of a forest stand from a digital surface model C to a plurality of cells Ci,jCombinations of (a) and (b).
From each grid cell ci,jThe z-value corresponding to the cell at which the highest z-value scan point was calculated in the interior ground LiDAR data is recorded as the z-value of that cellAt the same time, the point clouds in the vertical direction within each cell are stored in the corresponding grid cell. And secondly, detecting the crown vertex from the DSM by using a local maximum algorithm and adopting a size self-adaptive smooth window based on the tree height. And thirdly, pouring water into the whole tree crown forest stand, and synchronously spreading the water in each tree crown to draw the boundary of the tree. And fourthly, optimizing and judging the accuracy of the crown boundary by combining the gradient direction on the DSM and the height difference between the crown boundary unit and the two nearest crown vertexes. Fifthly, the estimation result of the boundary of the canopy is compared with the actually measured data, and the effectiveness of the algorithm is verified.
The technique for extracting individual tree profiles from airborne lidar data is set forth with reference to the workflow shown in fig. 1. Firstly, converting LiDAR data scanned in a research area into DSM (namely C) by adopting a rasterization method; second, the local maximum algorithm is applied to the DSM to detect crown vertices (i.e., cells c with local maximum z-value)i,j) (ii) a Thirdly, positioning the boundary of each crown by adopting a water pouring method for water spreading in each crown; optimizing and guiding the division of the tree crown boundary by utilizing the gradient direction and the height difference; fifthly, the crown segmentation is finished.
2.2 laser data
Lidar data acquisition was performed using a Riegl LMS-Q680i laser scanner at 750 meters altitude and 180 km/h flight speed with 65% side circles for one flight path. The sensor records the echo signal of the laser pulses with a time interval of 3 ns. The scanner emits a 1550nm wavelength at 300kHz pulse repetition frequency with a scan frequency of 80Hz, a scan angle of + -30 DEG, and a field of view of 60 deg. The beam divergence was 0.5mrad and the beam footprint size was 37.5 cm. The average distance between the ground scanning points is 0.45m, and the pulse density is about 9.58 points/m2. The final extracted point cloud is stored in LAS 1.2 format.
2.3 artificial measurement data of sample plot
The field data in the area under study were measured manually in the field and a total of 12 square plots (side length 20 m) were created at the data acquisition points based on forest type, tree growth density and complex terrain. All the samples are classified into eucalyptus (n-6) and fir (n-6) according to the tree species composition. The center of the field plot is identified using Trimble GeoXH6000 Global Positioning System (GPS) (Trimble, Sunnyvale, Calif., USA), and these points are corrected using high precision real-time differential signals received by Jiangsu continuous operating reference stations (JSCORS), thereby achieving sub-meter accuracy. The study measured heights greater than 2 meters and used a vertex V altimeter: (Sweden) measured crown apex height. The crown width is obtained from the average of two values measured in two perpendicular directions from the crown apex position. These plots can be classified into three categories, low growth density (plots 1,4, 10 and 11. low density trees ranging from 0 to 25 per plot), medium growth density (plots 2, 5, 7, 8 and 12. medium density trees ranging from 25 to 35 per plot), and high growth density (plots 3,6 and 9. high density trees ranging from more than 35 per plot). In addition, in recent decades, tree thinning has occurred in fir research areas, resulting in mixed growth of various broad-leaved tree species such as camphor, acacia, elm, magnolia, etc. Eucalyptus, as a fast-growing tree species, has strong soil nutrient absorption capacity, thus leading to the death of other vegetation in the area and constituting a pure artificial forest which is not interfered by other tree species.
2.4 DSM Generation and crown vertex detection
First, point cloud data acquired by an airborne laser radar is classified into ground points and above-ground points (scanned tree points) by a morphological filter. Then, each cell c is usedi,jThe maximum height of all scan tree points within generates a DSM with a grid cell size d. Cell c without any point cloudi,jGenerated by the height interpolation of its neighboring cells, where i, j denotes ci,jList of i row and j column in generated cell set C (i.e., DSM)And (4) cell.
In the last decade, the tree thinning phenomenon occurs in the fir forest research area. Many broadleaf trees are grown in these woodlands intermixed with fir wood, resulting in inconsistent forest canopy composition and crown shape characterization. Broad-leaved crowns have many wild branches, which makes detection of the crown apex difficult. Therefore, the DSM is smoothed by adopting a variable-scale Gaussian kernel smoothing filter, and the interference generated by mixed growth of other different broad-leaved tree species in the current plot is eliminated. According to the priori knowledge of the growth characteristics of the fir and the eucalyptus, the higher the fir and the eucalyptus tree body is, the larger the crown size is. Therefore, we analyze the height of the DSM, and use the DSM height as an important indicator for selecting the size of the smoothing filter window. For forest plots with larger average tree heights, the corresponding forest plots are morphologically filtered using a smoothing filter with a larger window size, and vice versa. Here, the window size of the smoothing filter is set to 3 × 3 cell size for plots with smaller tree heights; the 5x5 cell size is set for plots with larger tree heights. After our DSM smoothing, the local maximum finding algorithm is used to identify the location of each crown vertex in the filtered DSM and to use it as a seed point for the flood-run algorithm to achieve crown segmentation.
2.5. Concept of water spread and energy function control
2.5.1 initial segmentation based on the idea of pouring Water
In order to accurately extract boundary parameters of each tree, an algorithm based on a water pouring idea is researched, and spatial distribution of a single tree is obtained from DSM. According to the concept of pouring water into the pit (inverted crown profile), we synchronously control the water that begins to flow into the pit according to the height of the tree. The schematic diagram of fig. 2 shows the detailed spreading process of the water flowing into each pit. To facilitate our algorithm implementation using computer programming, the inverted DSM of each forest plot is based on a height separation HintervalIs divided vertically into s layers and during pouring water enters the lowest layer first until the lowest layer is filled with water. When the water level rises, water spreads to the adjacent layer (upper layer). In this order, the pouring process continues until it is at its maximumThe upper layer is filled with water until it is full. The water first enters the upper part of the forest canopy, i.e. the lower layer of the pit, with less overlap between the crowns in the upper part of the canopy, and the individual canopy profiles can be clearly delineated by the edges of the water in each pit. As the water pouring process progresses, the recorded upper crown contour and lower layer morphological information of each pit will be inherited, thereby guiding the segmentation of the upper crown overlap region (fig. 2).
Figure 2 illustrates the detailed steps of the algorithm. Step one, inverting DSMs of trees, regarding each spatial crown as a pit, and layering the inverted DSMs according to height intervals. And step two, pouring water, wherein the water naturally flows to the lowest layer of the pit. Step three: and the water pouring process is continued to cause the water level to rise and the area to synchronously increase, so that the edges of all the crowns are divided, and the upper intersection areas of the adjacent crowns are accurately separated by taking the edge information corresponding to the lower layer information filled with water as guidance. And step five, finishing the water pouring process and accurately describing the spatial contour of each tree.
The expansion process of water spread in the inverted tree model to accurately extract individual tree crowns is described with reference to the schematic diagram of the algorithm shown in fig. 2. (a) Take a side view of three trees of different species as an example. (b) The side view of the tree is inverted 180 degrees, looking like three pits. (c) Water was poured into the three wells and the boundaries of the crown were delineated by boundaries of water. (d) And (e) a first step of layering the DSM according to a given height interval and pouring water into all the pits (inverted crowns), the water flowing first into the lowest layer of the pits. (f) And (g) as the water pouring process continues, the water level rises in each crown and increases with the area synchronously, so as to divide the edges of the crowns, and the intersection area of the upper parts of the adjacent crowns is accurately separated according to the edge information corresponding to the lower layer information filled with water as guidance.
Our water-spreading process is actually operated on DSM and the mathematical expression of our process is set out below. After highly partitioning DSMs into different levels, region growing begins with the highest detected level l1The crown vertices of (a) located in cell C of the digital surface model Ci,jIn (c) thati,j＝＝tk，k∈[1,2,...r]&tk∈l1. And then combining the phenotypic characteristics of the shape of the crown, namely, the z value of the scanning points on the surface of the crown is gradually reduced from the same polarity at the top of the crown to each direction, and simulating water spreading by adopting a neighborhood search strategy. When the crown vertex t is detectednAnd its location on DSM, to find the current layer l1Each of which is connected with ci,jAdjacent cell ofAnd satisfies the phenotypic characteristics of crown shapeTo help determine the course of water flow. The water region of each crown then continues to spread according to the crown-shaped phenotypic characteristics, fromStart and synchronously creep to wherein To representAnd satisfiesNext, the iterative process is repeated until all the crown vertices c are matchedi,jHaving a partnership and belonging to the current layer l1The cells of (a) are all spread by water ((a 1), (a2) in fig. 3).
At a height interval of l1After the spread of the water has been completed, the corresponding area on the DSM where each pit is submerged will be inherited to the next height interval l2In (1). Then, at a height interval of l2The crown with the newly added crown vertex is added into the area growth, the water level growth is not completed, and the crown vertex exists in the upper layer l1Will continue to growTheir area grows and the overlapping area between the two crowns will be correctly divided according to the water spread cases inherited from the upper layers. The water spreading pattern in each pit will be applied in height order to the next height interval until all height layers have been treated, as shown in particular in figure 3.
The process of water spread growth in the corresponding DSM and crown is illustrated with reference to the program run diagram shown in fig. 3. With the pouring of water synchronized to each crown from highest to lowest, (a1), (b1), (c1), (d1), (e1), and (f1) represent side views of the LiDAR data. (a2) (b2), (c2), (d2), (e2) and (f2) are water-expanded top views on DSM in synchronization with the pouring processes shown in (a1), (b1), (c1), (d1), (e1) and (f1), respectively. The lidar data and the corresponding DSM area expanded by the water level are set to the same gray level. The bold lines represent the boundaries of the crown submerged in water at each step of the algorithm.
2.5.2 control guidelines for Water spreading sequences of various crowns
It is impossible to really realize the water synchronous spreading in each crown by using a computer program. The computer can only treat the water spread in each crown in a random order in the same height interval. This uncertain order can lead to false water level propagation when adjacent trees have overlapping crown areas in the same height interval. For example, in a vine delay in a plurality of crowns, the intersection area originally belonging to crown a is occupied by the adjacent crown B, since the randomness of the computer program gives priority to the water spreading operation to crown B. Therefore, a standard is urgently needed to control the water spreading sequence in each crown and provide a standard to evaluate the rationality of the sequence. Therefore, we combine the gradient direction information of the DSM with the height difference information of each unit on the DSM, and constitute the control criteria for water spreading order management. The details are as follows.
Represents the calculated gradient vector for each cell on the DSM,indicating that the vector to the north is measured in a clockwise mannerThe included angle of (a). The energy function is then designed to evaluate the correctness of the water spread order in each crown, as shown below.
Wherein Q denotes the current boundary element c belonging to each crown in each step of water spreadi,jThe number of the cells.Representing the vertex t from the crownkA vector pointing north.Representing vectorsTo vectorClockwise angle of (c). When the boundary cell ci,jIt has a real crown vertex tkCrown of tree, boundary unit ci,jAngle value ofApproximately equal to the calculated gradient correlation angleThus, correct water spread in each crown will result in the equationA minimum value is reached. Conversely, when an improper water spread sequence results in an incorrect crown boundary division, from the wrong crown vertex tk+1CalculatedAnd the current boundary ci,jIs not uniform in the gradient direction, which results inThe value of (a) becomes larger.
In addition, the optimal segmentation result of the crown is guided by using the height difference in the energy function. According to the idea, the height difference from a point on the mountain valley route between two mountains to the highest vertex of the two mountains should be the largest. Thus, cell c on the boundary of the crown due to water spreadi,jHeight difference between two adjacent nearest crown vertexesIt will gradually increase as the water spreads. Height difference between unit and adjacent crown vertex on real boundary of crown formed by water spreadingA maximum value should be reached. Therefore, the second term on the right side of equation (3) will gradually decrease with water spread, and according to the correct segmentation process, in the water spread process of finding the real boundary of each crown, the value of equation (3) for each crown contour should continuously decrease until reaching the minimum value, which represents that the water spread reaches the real boundary of the crown. In contrast, the value of equation (3) increases during water creep, which suggests that there is a segmentation error in the cross-crown region, resulting in a backtracking of the procedure. The specific description is shown in fig. 4.
The control criteria designed to determine the crown boundaries are illustrated with reference to the schematic diagram shown in fig. 4. (a) Representing DSM gradientsAnd (5) calculating the direction. (b) Is a close-up of the rectangle in (a). When any unitBelongs to its real crown vertex tkWhen the crown is covered, the gradient directionShould be approximately equal to t from its real crown vertexkTo the current cellIn the direction of (i.e. of)And a unitIs a crown top point tkRather than the crown apex tk+1The crown of the tree, soThis destroys the downward trend of the energy function values and suggests a wrong water spread in the intersection between the crowns according to the rise of the energy function. (c) It is illustrated that the accuracy of crown segmentation (i.e. the cells c on the true boundary of the crown) is evaluated in connection with the height difference between the delineated crown boundary cells and the two nearest crown verticesi,jHeight difference from nearest two crown vertices to any other cell cmiddleThe corresponding height differences are all large). (d) The display uses the nearest neighbor principle to judge the two boundary cells closest to the crown vertex.
2.5.3 energy function directing crown segmentation
In order to make up the limitation of computer program on the synchronous water spreading of all the tree crowns, the water spreading sequence is regulated by using the constructed energy function value, and then the segmentation result of the cross tree crown region is optimized. The continuous drop in the value of the energy function in the water spread iteration step of each crown is reasonable, as described by equation (3) in section 2.5.2. If during water spread an unexpected increase of the energy function occurs, this means that the water in the respective crown spreads to the adjacent crown area. Therefore, the corresponding water spreading step of the tree is cancelled, the priority of water spreading is given to other subsequent crowns, the water spreading sequence of the current tree is adjusted, and the global optimization of the crown segmentation result is ensured. In plot 1 (eucalyptus), the water spread in 29 crowns is shown in fig. 5, where the boundaries of water spread in the 100 th, 500 th and 1168 th iteration steps are represented by different gray scales. Meanwhile, in the iterative process of the algorithm, the energy function value calculated according to the boundary divided by the water spread in each crown is shown in fig. 5, and the global optimality of water level expansion is ensured through the adjustment of the algorithm.
Referring to fig. 5, the spreading order of our water spreading algorithm on each crown is judged by the rising and falling trend of the constructed energy function value. (a) The water spread and water boundaries for each crown during the iteration are depicted with lines of different shades of gray. (b) And taking the descending trend of the energy function value of each crown as the sequence for evaluating the water spread in each crown and ensuring the correct segmentation result of the intersection area between the crowns.
2.6 evaluation of accuracy
And comparing the result obtained in the section 2.5 with the actually measured data, and verifying the accuracy of the algorithm. And calculating the precision comparison between the measured crown top point and the measured crown top point in the research area by using a formula.
r (call) is the detection rate of trees, p (precision) is the correctness of the detected tree, f is the overall precision of the detected tree, tp (true positive) is the number of correctly detected trees, fn (false negative) is the number of undetected trees (missing errors), fp (false positive) is the number of trees that are not present in the region but are erroneously added (error-scoring errors).
In order to verify the matching degree of the individual crown contour line and the artificial crown contour line, the accuracy of the crown contour line is verified by comparing the actual-measured crown radius with the crown radius detected by an algorithm. We have chosen the correctly detected crown for crown radius comparison. The accuracy of the crown radius and height is evaluated with R-squared, Root Mean Square Error (RMSE) and relative RMSE (i.e. the ratio of RMSE to the observed mean).
3. Results
Our algorithm was applied to 12 plots including fir forest and eucalyptus forest. Figure 6 shows the results of individual tree crowns for six selected plots based on DSM. The algorithm is evaluated, and the accuracy of the algorithm is verified. We considered three typical eucalyptus plots (plots 1-3) and three fir plots (plots 4-6) of different tree growth densities. FIG. 6 shows the position of a set of three square plots (1,2,3) of growing Eucalyptus trees of 20 meters in length; the other group is the location of three square plots (4,5,6) of growing fir wood of length 20 meters. Tree growth density (plants/m) of eucalyptus2) From low to high, 0.06 (plot 1), 0.09 (plot 2) and 0.17 (plot 3) are respectively arranged. Density of trees grown in China fir sample plot (plants/m)2) From low to high, 0.06 (plot 4), 0.07 (plot 5) and 0.09 (plot 6) are arranged in sequence. Subsequently, 12 plots (including the selected 6 plots) were used for accuracy evaluation. Table 1 shows the accuracy of our algorithm for these 12 graphs. The algorithm has good performance, and the average detection rate (r) is 0.93; further, the highest accuracy (f) of the 12 graphs was 0.92. In the Eucalyptus group (plots 1-3,7-9), plot 2 had the greatest r value (0.97), plot 3 had the greatest p value (0.89), and plot f had the greatest f value (0.92). In fir wood (plots 4-6,10-12), plot 4 had the greatest r value (0.96), the greatest p value (0.73), and the greatest f value (0.83). As is apparent from table 1, the misclassification error is much larger than the missing error in the graph. The block 6 has a large misclassification error and a missing error. Due to the appearance in the forestMore serious needle-wide mixed and standing wood density is high, and the problem of excessive segmentation of the fir is more serious.
FIG. 7 shows the range of RMSE values for crown radii from 0.13m to 0.31 m. The RMSE average value of the eucalyptus plots is 0.24m, the rRMSE value range of the crown radius is 11.76-13.29%, and the average value is 12.21%. Plot 3(RMSE 0.17m, tree growth density 0.17, f 0.92) RMSE was the lowest, 11.83%. The RMSE value range of the fir crown width is 0.17-0.37 m, and the average value is 0.24 m. The range of the rmse value for the crown radius is 12.05% to 14.05%, with an average value of 13.56%. Plot 6(RMSE 0.19m, tree growth density 0.09, f 0.75) RMSE was the lowest, 13.01%. In all plots, the RMSE was less than 0.37m at the highest and the rRMSE was less than 14.05% at the highest. The result shows that the algorithm has stronger crown outline delineation capability. The average rRMSE value of fir wood is higher than that of eucalyptus wood. The result shows that the algorithm has certain advantages in the aspect of the precision of the eucalyptus canopy. FIG. 8 shows that the RMSE values for eucalyptus plot tree heights range from 0.82m to 1.71m, with an average value of 1.01m, and the RMSE values for tree heights range from 5.65% to 6.81%, with an average value of 6.30%. Plot 9(RMSE 0.92m, tree growth density 0.12, f 0.88) RMSE was the lowest, 5.65%. The RMSE value of the crown width of the fir is between 0.87m and 0.94m, and the average value is 0.89 m. The tree height rRMSE values ranged from 4.28% to 6.96%, with an average of 5.83%. Plot 11(RMSE 0.88m, tree growth density 0.06, f 0.81%) RMSE was the lowest, 4.28%. In all plots, the RMSE was up to 2.46m, and the rRMSE was up to 10.58%. The result shows that the algorithm can accurately obtain the tree heights of the eucalyptus and the fir. The data show that the difference in rmse values for the two tree species is only 1% in tree height. The result shows that the algorithm has certain universality in the aspect of tree height extraction.
TABLE 1 evaluation of precision of Tree segmentation in test plots
TP number of correctly detected trees. FP the number of additional trees not present in the region (error separation error) FN the number of undetected trees (missing errors). r is the detection rate of trees. p, correctness of the detected tree. f, the overall precision of the detected tree.
See FIG. 6 for a tree crown test using our algorithm for six selected plots. (a) The terms (b), (c), (d), (e) and (f) represent the individual crown divisions of the plots of Nos. 1,2,3, 4,5 and 6, respectively. The classification was done by tree growth density of eucalyptus (plots 1,2,3) and fir (plots 4,5,6), with black (grey) borders indicating detected crowns. The black dots represent the manually measured crown vertices and the black crosses represent the crown vertices detected with our algorithm.
Referring to fig. 7, the field measured crown radius values are compared to the crown radius values detected in the 12 blocks of regions by our algorithm. (a), (b), (c), (d), (f), (g), (h), (i), (j), (k) and (l) represent crown radius comparison plots for plots 1,2,3, 4,5,6, 7, 8, 9, 10, 11 and 12, respectively.
Referring to fig. 8, the measured in-situ tree height is compared to the tree height detected by our algorithm for estimating 12 plots. (a) And (b), (c), (d), (e), (f), (g), (h), (i), (j), (k) and (l) represent comparative tree heights of 1,2,3, 4,5,6, 7, 8, 9, 10, 11 and 12 patterns, respectively.
4. Discussion of the related Art
4.1 Algorithm efficiency analysis
First, we analyzed the detection rate of crown vertices in the study area. In fig. 6, we have divided the selected 6 plots into three categories based on tree growth density, including low growth density (plots 1,4), medium growth density (plots 2, 5), high growth density (plots 3, 6). For eucalyptus, the overall accuracy of the detected tree (f) is highest, and the tree growth density of the plot is also high (0.17 trees/m)2). As the density of tree growth increases, the detection precision tends to increase. This is because eucalyptus plots are uniformly planted and have a certain regularity. Meanwhile, the growing density of the plot trees is increased, and smaller smooth windows are selected from uniform plotsThe mouth has better smoothing effect on DSM shape. And the measured crown vertex precision is reduced along with the increase of the growth density of the fir tree. The result shows that the algorithm has a good segmentation effect on the low-tree growth density land of the fir. In plots 4 and 6, when the tree growth density increased by 0.03 plants/m2When the time is long, the total precision (f) of the fir detected by the land is reduced by 0.08. This result indicates that the overall accuracy of the trees detected in the graph is more sensitive to the density of tree growth. When the tree growth density is too high (land parcel 6 is 0.09/m)2) Over-segmentation is more severe, which is a common error when the density of fir growth is high. Due to the difference of characteristics and human factors of the two tree species of the fir and the eucalyptus, the overall precision (f) of the detection result of the fir is lower than that of the eucalyptus. Eucalyptus has strong soil fertility competitiveness and is an economic tree species. Every few years, eucalyptus is cut down to gain economic value and new saplings are planted; thus, the lower layers of eucalyptus have little other plants. China fir has weak competitive factors on forest trees, and some secondary plants also grow under the China fir; therefore, the growing environment of fir wood is more complicated than that of eucalyptus under the same climatic conditions. The above causes the overall precision of the fir wood to be lower than that of the eucalyptus. The result shows that (1) the tree detection precision of the southern China fir is reduced along with the increase of the growth density of the tree; (2) the detection precision of the eucalyptus increases along with the increase of the growing density of the tree; (3) the detection precision of the fir is lower than that of eucalyptus.
Next, we analyzed the accuracy of the crown radius. The crown radius accuracy of our algorithm detection increases with the density of growing trees of the same species. In fig. 7 and 8, we analyzed the crown radius and the tree height and compared them with the field measurements in plots 1-6. The accuracy of the eucalyptus and fir crown radius values increases as the density of the tree growth increases. In eucalyptus and fir, the rmse values for the high and low growth densities were 1.46% (plots 1, 3) and 1.05% (plots 4, 6), respectively. The results show that an increase in the density of tree growth has an effect on the accuracy of the crown radius. In plots 1 and 4, the densities of the two plots were similar, with the canopy radius accuracy of eucalyptus being 0.76% higher than that of fir. This is because the shape of the fir canopy varies more than that of the eucalyptus canopy, increasing the difficulty of calculating the canopy radius. Due to the high calculation difficulty, the precision of the eucalyptus tree under the same tree growth density is slightly higher than that of the fir tree. In our algorithm, we segment the crown according to the correctly detected crown vertices. At higher tree growth densities, the crowns of the two species are more uniform (with fewer unique shapes), reducing the impact of secondary trees. Under the condition, each height interval of each crown is searched by using a synchronous search mode, and the algorithm effect is good. The synchronous water spreading ensures that the boundary area of each crown is more regular and the boundary of the crown is more accurately determined.
For the tree heights of six land parcels, the detection of the tree heights by all the land parcels has higher precision. The results show that the change in tree growth density has no significant effect on the accuracy of estimating the tree height. After crown vertex detection, our algorithm shows that the tree height is independent of the accuracy of crown radius detection. The rmse values for all plot heights were below 8.75%. The result shows that the algorithm has good effect of obtaining the tree height, and the growth density of the tree species and the plot tree has no obvious influence on the tree height obtaining.
4.3 height Interval HintervalInfluence on algorithm
In addition to smooth window and point cloud density, height interval is a major parameter that affects the accuracy of the algorithm. The height interval may control the accuracy of the boundary and the steps of the search process during the search process. Each control of water spread is based on a height interval. As the height interval decreases, the steps performed in the program increase, thereby increasing the computation time of the algorithm. To further investigate the effect of fir and eucalyptus on height separation, we adjusted the height separation threshold from 0.2 meters to 2 meters, with each change of 0.2 meters, and analyzed changes in the crown radii and calculated time rmse values for eucalyptus (plots 1-3) and fir (plots 4-6). FIG. 9 shows the effect of height separation on the value of crown radius rRMSE. In fig. 9, as the height interval increases, the rmse value of the crown radius of eucalyptus and fir test continues to increase. The smaller height interval can distinguish the vertex positions of the crowns belonging to different height layers, thereby ensuring better synchronous water level spreading. The larger height interval causes the search result of the region growing to have large data quantity at each step, much interference and reduced overall precision. Fig. 9 illustrates that setting the height interval to 0.2m results in the lowest rmse values for eucalyptus and fir. Fig. 9 shows that a small height spacing increases the accuracy of the detection of the crown radius. The rRMSE value distribution of the fir crown radius is 0.5 percent higher than that of eucalyptus according to the change of the algorithm height interval adjustment quantity. Therefore, the rmse of fir is more sensitive than eucalyptus in high intervals.
The calculation time of eucalyptus and fir rapidly decreases as the height interval increases, as shown in (c), (d) of fig. 9. The height difference of the study area ranges from 20 to 40m, and the height interval is set to 0.2m here for better crown radius accuracy. While lower height spacing can improve the accuracy of fir and eucalyptus crown radii, using our algorithm for larger woodland areas requires more computation time. The selection of the height interval has an important influence on the accuracy of crown radius estimation and the calculation time.
See figure 9 for the effect of height separation on crown radius rmse and computation time. (a) Eucalyptus measured distributions of crown radius rRMSE scores as a function of height separation. (b) The distribution of the measured radius rmse of the fir crown as a function of height separation. (c) distribution of individual eucalyptus isolates calculated time as a function of height interval. (d) The distribution of the calculation time of the isolation of the individual fir plants varies with the height interval.
5. Conclusion
The accurate single plant separation algorithm can effectively improve the sustainable accurate management quality of the forest. Our method is mainly divided into three parts. Firstly, extracting tree top information from airborne laser point cloud through a local maximum algorithm and Gaussian kernel smooth filtering. Second, we simulated the physical principle of pouring water into the inverted tree surface model to achieve the segmentation of the initial crown. Thirdly, an energy function is constructed by utilizing the gradient direction and the height difference on the DSM to assist in determining the tree crown boundary, so that the problem that the tree crown overlapping area cannot be accurately divided is optimized, and the defect that water spreads in each tree crown synchronously is overcome. And finally, comparing the calculated result with the actual measurement result of the subtropical research area of the southern peak forest land. The result shows that the algorithm detects and sums up on the top of eucalyptus and firThe crown radius detection method has high precision. The accuracy of the detection of the eucalyptus tree top is high (recall is 0.94, precision is 0.81, and f is 0.86), and the next time of the cedar (recall is 0.90, precision is 0.71, and f is 0.80). Meanwhile, the average precision (R) of the fir crown radius20.79 RMSE 0.25m, rRMSE 13.56%) is lower than that of eucalyptus (R)20.77, RMSE 0.24m, RMSE 12.21%). The trees of fir and eucalyptus have similar high precision (less than 1%). Meanwhile, the sensitivity of the fir (with the range of the crown radius rmse value being 3.30%) to the height interval is higher than that of the eucalyptus (with the range of the crown radius rmse being 2.73%), which causes more needles and broad mixing phenomenon due to the fir-like intermediate cutting forest strategy and brings more interference. But the overall precision achieved by the method meets the requirement of forest stand fine measurement.
Claims (4)
1. The individual tree crown segmentation method for water pouring spread and energy function control of airborne laser point cloud is characterized by comprising the following steps: it comprises the following steps:
a. converting tree point cloud data acquired by an airborne laser radar into a digital surface model C; rasterizing the digital surface model C of the forest stand to convert the digital surface model C into a plurality of cells Ci,jA combination of (1); then, a local maximum algorithm is utilized to search a unit C corresponding to the local maximum of the z value from the digital surface model Ci,jFurther detecting the crown top point;
b. and taking the crown top point as a seed point of a water pouring spreading algorithm, pouring water into the whole crown forest stand, and synchronously spreading the water in each crown to draw the boundary of the tree.
2. The method for segmenting the crown of the single plant as claimed in claim 1, wherein: the specific process of the step b is as follows:
inverting the digital surface model C of the tree, regarding each spatial crown as a pit, and layering the inverted digital surface model C according to the height interval;
pouring water, wherein the water naturally flows to the lowest layer of the pit;
continuing pouring water to cause the water level to rise and the area to grow synchronously, starting spreading of the water, dividing the edges of all the crowns, and taking edge information corresponding to lower-layer information filled with the water as guidance to accurately separate the upper intersection areas of the adjacent crowns;
and (5) finishing the water pouring process and describing the spatial outline of each tree.
3. The method for segmenting the crown of the single plant as claimed in claim 2, wherein: the specific process of water spread is as follows:
after the digital surface model C is divided into different levels by height, region growing starts at the highest level l detected1Crown vertex t of treekThese crown vertices tkCell C located on digital surface model Ci,jIn (c) thati,j＝＝tk，k∈[1,2,...r]&tk∈l1(ii) a Then combining the phenotypic characteristics of the shape of the crown, namely, the z value of the scanning laser point on the surface of the crown is gradually reduced from the same polarity at the top of the crown to each direction, and simulating water spreading by adopting a neighborhood search strategy; when the crown vertex t is detectedkAnd its location on DSM, to find the current layer l1Each of the interior and the unit c as the crown topi,jAdjacent cell ofAnd satisfies the phenotypic characteristics of crown shapeTo help determine the course of water flow; the water region of each crown then continues to spread according to the crown-shaped phenotypic characteristics, fromStart and synchronously creep to wherein To representAnd satisfiesNext, the iterative process is repeated until all the cells c associated with the crown verticesi,jUnits with a partnership and belonging to the current level l1All the cells are spread by water;
at a height interval of l1After the spread of the water inside is completed, the corresponding area of the digital surface model C where each pit is immersed in the water will be inherited to the next height interval l2Performing the following steps; then, at a height interval of l2The crown with the newly added crown vertex is added into the area growth, the water level growth is not completed, and the crown vertex exists in the upper layer l1Will continue their area growth and the overlapping area between the two crowns will be correctly segmented according to the water spread inherited from the upper layers; the water spreading pattern in each pit will be applied in height order to the next height interval until all height layers have been treated.
4. The method for segmenting the crown of the single plant as claimed in claim 1, wherein:
combining the gradient direction information of the digital surface model C with the height difference information of each unit in the digital surface model C to control the water spreading sequence, namely:
design energy function Boundarryenergy (t)k) To evaluate the correctness of the water spreading sequence in each crown, Boundaryenergy (t) during the water spreading processk) The value should continue to decrease until a minimum value is reached;
wherein ,
representing the calculated gradient vector for each cell in the digital surface model C,
indicating that the vector to the north is measured in a clockwise mannerThe included angle of (A);
q denotes the current border element c belonging to each crown in each step of the water spreadi,jThe number of (2);
representing the vertex t from the crownkA vector pointing north;
representing vectorsTo vectorClockwise angle of (d);
representing the vertex t from the crownkTo the current cell ci,jA vector of (a); beta is a weight coefficient.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910632238.9A CN110598707B (en) | 2019-07-13 | 2019-07-13 | Single plant tree crown segmentation method for airborne laser point cloud water pouring spreading and energy function control |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910632238.9A CN110598707B (en) | 2019-07-13 | 2019-07-13 | Single plant tree crown segmentation method for airborne laser point cloud water pouring spreading and energy function control |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110598707A true CN110598707A (en) | 2019-12-20 |
| CN110598707B CN110598707B (en) | 2023-06-06 |
Family
ID=68852794
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910632238.9A Active CN110598707B (en) | 2019-07-13 | 2019-07-13 | Single plant tree crown segmentation method for airborne laser point cloud water pouring spreading and energy function control |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110598707B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2022067598A1 (en) * | 2020-09-30 | 2022-04-07 | Nanjing Maoting Information Technology Co., Ltd. | Method of individual tree crown segmentation from airborne lidar data using novel gaussian filter and energy function minimization |
| CN115131678A (en) * | 2022-07-01 | 2022-09-30 | 杭州万林数链科技服务有限公司 | Block chain architecture construction method and system for digital intelligent forestry |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2009022278A (en) * | 2007-06-19 | 2009-02-05 | Tokyo Electric Power Co Inc:The | Method for recognizing treetop, device for recognizing treetop, and program for recognizing treetop |
| CN107832681A (en) * | 2017-10-16 | 2018-03-23 | 福州大学 | The high evaluation method of forest list ebon of joint LiDAR point cloud and synchronous remote sensing image |
-
2019
- 2019-07-13 CN CN201910632238.9A patent/CN110598707B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2009022278A (en) * | 2007-06-19 | 2009-02-05 | Tokyo Electric Power Co Inc:The | Method for recognizing treetop, device for recognizing treetop, and program for recognizing treetop |
| CN107832681A (en) * | 2017-10-16 | 2018-03-23 | 福州大学 | The high evaluation method of forest list ebon of joint LiDAR point cloud and synchronous remote sensing image |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2022067598A1 (en) * | 2020-09-30 | 2022-04-07 | Nanjing Maoting Information Technology Co., Ltd. | Method of individual tree crown segmentation from airborne lidar data using novel gaussian filter and energy function minimization |
| CN115131678A (en) * | 2022-07-01 | 2022-09-30 | 杭州万林数链科技服务有限公司 | Block chain architecture construction method and system for digital intelligent forestry |
| CN115131678B (en) * | 2022-07-01 | 2023-03-07 | 杭州万林数链科技服务有限公司 | Block chain architecture construction method and system for digital intelligent forestry |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110598707B (en) | 2023-06-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107063197B (en) | Reservoir characteristic curve extraction method based on spatial information technology | |
| Kraus et al. | Advanced DTM generation from LIDAR data | |
| Allouis et al. | Stem volume and above-ground biomass estimation of individual pine trees from LiDAR data: Contribution of full-waveform signals | |
| Lefsky et al. | Volume estimates of trees with complex architecture from terrestrial laser scanning | |
| CN110132373A (en) | A kind of large-scale inland lake and wetland water level inversion method based on Altimetry Data | |
| CN107644452A (en) | Airborne LiDAR point cloud roof dough sheet dividing method and system | |
| Shaw et al. | Comparison of UAV Lidar and imagery for beach monitoring | |
| CN107330898A (en) | Altitudinal vegetation zone quantitatively delineates computational methods and system | |
| Bretar et al. | Terrain surfaces and 3-D landcover classification from small footprint full-waveform lidar data: application to badlands | |
| CN111091079A (en) | TLS-based method for measuring dominant single plant structural parameters of vegetation in alpine and fragile regions | |
| CN110598707B (en) | Single plant tree crown segmentation method for airborne laser point cloud water pouring spreading and energy function control | |
| Esposito et al. | Evaluation of geomorphic changes and retreat rates of a coastal pyroclastic cliff in the Campi Flegrei volcanic district, southern Italy | |
| CN108520271A (en) | Submarine geomorphy type sorter design method based on factorial analysis | |
| CN113834547B (en) | River virtual station water level time sequence reconstruction method and system | |
| CN101464148B (en) | Three-dimensional image detecting, compiling and reconstructing system | |
| CN106447510A (en) | Terrace field information extracting method and terrace field information extracting system | |
| Zlatanović et al. | Comparison of an automated and manual method for calculating storm runoff response in ungauged catchments in Serbia | |
| CN114089371A (en) | Method for estimating underground soil leakage amount in karst region by using laser Lidar technology | |
| Peng et al. | Error assessment in two LiDAR-derived TIN datasets | |
| JP2006003332A (en) | Estimating method of ground level | |
| Yu et al. | Detection and volume estimation of mining subsidence based on multi-temporal LiDAR data | |
| Kim et al. | Detecting Individual Tree Position and Height Using Airborne LiDAR Data in Chollipo Arboretum, South Korea. | |
| Sato et al. | Airborne LIDAR data measurement and landform classification mapping in Tomari-no-tai landslide area, Shirakami Mountains, Japan | |
| Jamru | Correction pit free canopy height model derived from LiDAR data for the broad leaf tropical forest | |
| Caris et al. | Analysis of Altimetric Data Obtained by Lidar in an Area of Atlantic Forest in Southeastern Brazil |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |