CN113009481A - Forest surface feature imaging inversion method based on interferometric SAR radar - Google Patents
- ️Tue Jun 22 2021
Info
-
Publication number
- CN113009481A CN113009481A CN202110057861.3A CN202110057861A CN113009481A CN 113009481 A CN113009481 A CN 113009481A CN 202110057861 A CN202110057861 A CN 202110057861A CN 113009481 A CN113009481 A CN 113009481A Authority
- CN
- China Prior art keywords
- point cloud
- tree
- height
- point
- ground Prior art date
- 2021-01-15 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
- 238000000034 method Methods 0.000 title claims abstract description 27
- 238000003384 imaging method Methods 0.000 title claims abstract description 9
- 230000001788 irregular Effects 0.000 claims abstract description 20
- 238000001914 filtration Methods 0.000 claims abstract description 19
- 230000011218 segmentation Effects 0.000 claims description 28
- 238000000605 extraction Methods 0.000 claims description 20
- 241000196324 Embryophyta Species 0.000 claims description 19
- 238000004364 calculation method Methods 0.000 claims description 15
- 238000011835 investigation Methods 0.000 claims description 8
- 238000001514 detection method Methods 0.000 claims description 7
- 238000003064 k means clustering Methods 0.000 claims description 7
- 238000010606 normalization Methods 0.000 claims description 7
- 238000012545 processing Methods 0.000 claims description 7
- 239000002023 wood Substances 0.000 claims description 7
- 238000010276 construction Methods 0.000 claims description 6
- 238000012417 linear regression Methods 0.000 claims description 3
- 238000006748 scratching Methods 0.000 claims description 3
- 230000002393 scratching effect Effects 0.000 claims description 3
- 238000012360 testing method Methods 0.000 claims description 3
- 238000012549 training Methods 0.000 claims description 3
- 238000012795 verification Methods 0.000 claims description 3
- 230000009286 beneficial effect Effects 0.000 abstract description 4
- 238000012544 monitoring process Methods 0.000 description 9
- 238000005516 engineering process Methods 0.000 description 5
- 238000009825 accumulation Methods 0.000 description 3
- 238000011161 development Methods 0.000 description 3
- 230000018109 developmental process Effects 0.000 description 3
- 238000005259 measurement Methods 0.000 description 3
- 239000002028 Biomass Substances 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 241000894007 species Species 0.000 description 2
- 238000004177 carbon cycle Methods 0.000 description 1
- 230000008034 disappearance Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000035515 penetration Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/88—Radar or analogous systems specially adapted for specific applications
- G01S13/89—Radar or analogous systems specially adapted for specific applications for mapping or imaging
- G01S13/90—Radar or analogous systems specially adapted for specific applications for mapping or imaging using synthetic aperture techniques, e.g. synthetic aperture radar [SAR] techniques
- G01S13/9021—SAR image post-processing techniques
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/88—Radar or analogous systems specially adapted for specific applications
- G01S13/89—Radar or analogous systems specially adapted for specific applications for mapping or imaging
- G01S13/90—Radar or analogous systems specially adapted for specific applications for mapping or imaging using synthetic aperture techniques, e.g. synthetic aperture radar [SAR] techniques
- G01S13/9021—SAR image post-processing techniques
- G01S13/9023—SAR image post-processing techniques combined with interferometric techniques
Landscapes
- Engineering & Computer Science (AREA)
- Remote Sensing (AREA)
- Radar, Positioning & Navigation (AREA)
- Physics & Mathematics (AREA)
- Electromagnetism (AREA)
- Computer Networks & Wireless Communication (AREA)
- General Physics & Mathematics (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a forest ground object imaging inversion method based on an interferometric SAR radar, which utilizes a point cloud filtering algorithm to extract topographic information, performs median filtering denoising on original point cloud data, takes an initial ground seed point as a seed point for constructing an initial irregular triangulation network, repeatedly judges the gradient and the altitude difference threshold of a point which is not scratched into the ground point cloud, and obtains a final ground point cloud when no new ground point is added into the triangulation network, thereby greatly improving the accuracy of ground point cloud judgment and improving the denoising efficiency; the inversion method can effectively reduce errors, is simple and feasible, is particularly suitable for inverting forest parameters, is closer to the actual height for the inversion result of the tree height, and can be more beneficial to the actual application of remote sensing images.
Description
Technical Field
The invention belongs to the technical field of SAR radar application, and particularly relates to a forest surface feature imaging inversion method based on an interferometric SAR radar.
Background
In recent decades, with the general attention of people on global climate change, carbon cycle research and human sustainable development problems, the forest ecosystem is highly valued by governments and scientists all over the world, and the current situation and the change rule of forest resources are mastered in time, so that the forest ecosystem is very important for the life of the ecosystem and the human.
For the monitoring of forest resources, the scale can be divided into the macroscopic monitoring of the national level and the fine monitoring of the mountain land. The traditional investigation method is based on sampling theory and mainly carried out by taking ground investigation as a main method. The traditional survey method has the main problems that the ground measurement workload is large, the updating period is long, and the forest resource survey results with uniform time and space-time continuity are difficult to obtain nationwide; with the continuous development of remote sensing forest resource investigation technology, key forest resource monitoring factors such as tree species/forest land types, forest heights, accumulation amounts, biomass and the like can be obtained by adopting the remote sensing technology.
The method for forest resource investigation by adopting remote sensing technology can be divided into three types from the sensor, namely optical remote sensing (multispectral and hyperspectral), laser radar and Synthetic Aperture Radar (SAR), wherein the laser radar has the highest precision, but is only suitable for forest resource investigation in small areas, and the cost of large-area application is higher. The optical remote sensing acquisition mainly comprises information of the surface of the forest canopy, and has advantages in forest land type classification, tree species identification and the like, but has limitations in quantitative estimation of forest height, accumulation and the like. The SAR has the all-weather observation capability and relatively longer wavelength, and has certain penetration capability on vegetation leaf clusters such as forests and the like, so that the remote sensing observation quantity more related to forest vertical structure parameters can be obtained.
Through the development of the last 30 years, the SAR forest resource monitoring technology has made a lot of progress in the aspects of forest land type classification and change detection, forest height inversion, forest accumulation/biomass estimation and the like.
Although SAR technology is rapidly developed, there are still disadvantages, such as: the SAR image data has large speckle noise, is more difficult to effectively remove than noise from other sources, and seriously interferes the extraction of the ground feature information and the application effect of the SAR image, and even can cause the disappearance of the information when the noise is serious.
Disclosure of Invention
The invention aims to provide a forest surface feature imaging inversion method based on an interferometric SAR, which solves the problems that SAR image data in the prior art is large in speckle noise, difficult to effectively remove and seriously interferes with the extraction of surface feature information and the application effect of an SAR image.
The technical scheme adopted by the invention is as follows:
the invention provides a forest surface feature imaging inversion method based on an interferometric Synthetic Aperture Radar (SAR), which comprises the following steps of:
step 1, acquiring a group of SAR image data of a target to be detected, extracting topographic information by using a point cloud filtering algorithm, and removing non-ground points from an original point cloud; an asymptotic irregular triangulation network encryption filtering algorithm is adopted, and the method specifically comprises the following steps:
1) carrying out median filtering denoising on the original point cloud data, removing noise point cloud, and avoiding influence on the algorithm of the subsequent steps;
2) meshing point clouds, namely setting the size of a grid, dividing the denoised point clouds into regular grids according to x and y coordinates of space points in the point clouds, and taking the lowest elevation point in each regular grid as an initial ground seed point;
3) taking the initial ground seed points as seed points for constructing an initial irregular triangulation network, introducing slope threshold value judgment in the process of constructing the initial irregular triangulation network, namely firstly performing slope calculation on adjacent network construction nodes when constructing the triangulation network, adding the nodes into the network construction if the slope threshold value is smaller than the slope threshold value, otherwise marking the nodes as non-ground points, and not participating in calculation in the later steps to finally obtain the initial irregular triangulation network;
4) on the basis of the initial irregular triangulation network, judging the gradient and the altitude difference threshold value of points which are not scratched into the ground point cloud, if the conditions are met, scratching into the ground points, and carrying out iterative encryption on the ground irregular triangulation network;
5) and repeating the step 4), stopping iteration when no new ground point is added into the triangulation network, and obtaining the triangulation network point cloud meeting the conditions, namely the final ground point cloud.
Step 2, single tree parameter extraction and forest stand parameter extraction, wherein the single tree parameters mainly comprise the height, the crown width and the position of a single tree, and the forest stand parameters mainly comprise the plant tree density and the average height of the forest stand;
2.1, extracting parameters of the single wood:
obtaining ground point clouds and non-ground point clouds contained in forest region point cloud data on the basis of point cloud filtering processing, wherein the forest region non-ground point clouds are basically equivalent to vegetation point clouds, performing elevation normalization processing on the original vegetation point clouds, determining the elevation of each space point in the output point clouds to be the absolute height of the point relative to the ground, then performing single-tree point cloud segmentation by using the normalized point clouds, and finally extracting the tree height and the crown width according to the single-tree point cloud segmentation result;
the method comprises the following steps of (1) carrying out single-tree point cloud segmentation by adopting a K-Means clustering point cloud segmentation algorithm based on normalized vegetation point cloud height layering:
1) point cloud normalization: and dividing vegetation point cloud data obtained by point cloud filtering into different grids according to the generated DEM image resolution size by taking the horizontal X, Y coordinate as the basis, and subtracting the value of the corresponding DEM grid from the elevation of each vegetation point falling in the grids to obtain the normalized vegetation point cloud, wherein the height of the midpoint of the vegetation point cloud is equivalent to the vegetation height.
2) Single-wood point cloud segmentation: dividing the vegetation point cloud into a plurality of layers according to the height distribution condition of the normalized vegetation point cloud, setting a neighborhood detection window size in each layer, carrying out height local maximum value detection, carrying out three-dimensional space point K-Means clustering on the vegetation point cloud of each layer by taking a local maximum value point as an initial clustering center, judging whether the horizontal distance between clustering centers in each clustering point cloud of adjacent upper and lower 2 layers is smaller than a set clustering center distance threshold value or not from the uppermost layer after the point cloud clustering of each layer is completed, merging the corresponding clustering point clouds in the upper and lower 2 layers if the horizontal distance is smaller than the threshold value until all the layers are completely compared and merged, and finally obtaining the single-tree segmentation point cloud;
3) parameter extraction: counting the highest point height of each single-tree point cloud to obtain single-tree segmentation point clouds, taking the highest point height as the extracted tree height, taking the horizontal coordinate of the point as a single-tree positioning coordinate, calculating the convex hull area of each single-tree segmentation point cloud projected onto the horizontal plane, taking the area as the nearly circular projection area of the tree crown, and calculating the diameter of a circle as the average crown width of the single tree;
2.2 forest stand parameter extraction
1) The plant and tree density is as follows: taking a square sample plot as an investigation area, counting forest plants and trees of the sample plot according to the sample plot single tree division result, and calculating the density value of the plants and trees of the sample plot by using a formula (1);
N＝n/S (1)
wherein N is the plant-tree density (plant/hm)2) (ii) a n is the number of tree plants in the sample plot; s is the area of the sample plot (hm)2)；
2) Average height of forest stand: the forest stand average height extraction method is characterized in that the correlation between the height of a quartile position on sample plot normalized vegetation point cloud data and the height of an actually measured tree is high;
firstly, selecting a certain amount of sample plot survey data as training samples, and calculating the height of the quartile on the sample plot normalized vegetation point cloud, wherein the specific calculation is as follows: ordering the normalized vegetation point clouds in the sample plot range according to the heights, then calculating the heights of upper quartiles of the total height, namely the heights of the upper quartiles of the normalized vegetation point clouds of the sample plot, establishing a linear regression equation between the heights of the upper quartiles of the normalized vegetation point clouds of the sample plot and the average height of the actually measured forest stand, and then carrying out precision verification by using reserved test samples; finally, realizing forest stand average height extraction of the sample plot to be detected according to the sample plot point cloud and a regression equation;
the actual measurement forest stand average tree height is calculated by adopting a cross-section area weighting calculation method, and the calculation formula is as follows:
wherein H is the average stand height HiIs the height of the ith tree, giThe chest height cross-sectional area of the ith tree is shown, and k is the number of forest stand plants.
The invention has the beneficial effects that:
(1) the method utilizes a point cloud filtering algorithm to extract topographic information, performs median filtering denoising on original point cloud data, can remove noise point clouds due to system errors or bird laser reflection points and the like, avoids the influence on the algorithm of the subsequent steps, takes the initial ground seed points as seed points for constructing an initial irregular triangular net, repeatedly judges the gradient and the altitude difference threshold value of points which are not scratched into the ground point clouds, and obtains the final ground point clouds when no new ground point is added into the triangular net, thereby greatly improving the accuracy of ground point cloud judgment and improving the denoising efficiency.
(2) Performing single-tree point cloud segmentation by using a K-Means clustering point cloud segmentation algorithm based on normalized vegetation point cloud height layering, performing elevation normalization processing on the original vegetation point cloud, wherein the elevation of each space point in the output point cloud is the absolute height of the point relative to the ground, performing single-tree point cloud segmentation by using the normalized point cloud, and finally obtaining the tree height and the crown width according to the single-tree point cloud segmentation result. The inversion method can effectively reduce errors, is simple and feasible, is particularly suitable for reversing forest parameters, is closer to the actual height for the inversion result of the tree height, and can be more beneficial to the actual application of remote sensing images.
Detailed Description
In order to enhance the understanding of the present invention, the present invention will be described in further detail with reference to the following examples. The following examples are only for illustrating the technical solutions of the present invention more clearly, and the protection scope of the present invention is not limited thereby.
Example 1
A forest surface feature imaging inversion method based on an interferometric SAR radar comprises the following steps:
step 1, acquiring a group of SAR image data of a target to be detected, extracting topographic information by using a point cloud filtering algorithm, and removing non-ground points from an original point cloud; an asymptotic irregular triangulation network encryption filtering algorithm is adopted, and the method specifically comprises the following steps:
1) carrying out median filtering denoising on the original point cloud data, removing noise point cloud, and avoiding influence on the algorithm of the subsequent steps;
2) meshing point clouds, namely setting the size of a grid, dividing the denoised point clouds into regular grids according to x and y coordinates of space points in the point clouds, and taking the lowest elevation point in each regular grid as an initial ground seed point;
3) taking the initial ground seed points as seed points for constructing an initial irregular triangulation network, introducing slope threshold value judgment in the process of constructing the initial irregular triangulation network, namely firstly performing slope calculation on adjacent network construction nodes when constructing the triangulation network, adding the nodes into the network construction if the slope threshold value is smaller than the slope threshold value, otherwise marking the nodes as non-ground points, and not participating in calculation in the later steps to finally obtain the initial irregular triangulation network;
4) on the basis of the initial irregular triangulation network, judging the gradient and the altitude difference threshold value of points which are not scratched into the ground point cloud, if the conditions are met, scratching into the ground points, and carrying out iterative encryption on the ground irregular triangulation network;
5) and repeating the step 4), stopping iteration when no new ground point is added into the triangulation network, and obtaining the triangulation network point cloud meeting the conditions, namely the final ground point cloud.
Step 2, single tree parameter extraction and forest stand parameter extraction, wherein the single tree parameters mainly comprise the height, the crown width and the position of a single tree, and the forest stand parameters mainly comprise the plant tree density and the average height of the forest stand;
2.1, extracting parameters of the single wood:
obtaining ground point clouds and non-ground point clouds contained in forest region point cloud data on the basis of point cloud filtering processing, wherein the forest region non-ground point clouds are basically equivalent to vegetation point clouds, performing elevation normalization processing on the original vegetation point clouds, determining the elevation of each space point in the output point clouds to be the absolute height of the point relative to the ground, then performing single-tree point cloud segmentation by using the normalized point clouds, and finally extracting the tree height and the crown width according to the single-tree point cloud segmentation result;
the method comprises the following steps of (1) carrying out single-tree point cloud segmentation by adopting a K-Means clustering point cloud segmentation algorithm based on normalized vegetation point cloud height layering:
1) point cloud normalization: and dividing vegetation point cloud data obtained by point cloud filtering into different grids according to the generated DEM image resolution size by taking the horizontal X, Y coordinate as the basis, and subtracting the value of the corresponding DEM grid from the elevation of each vegetation point falling in the grids to obtain the normalized vegetation point cloud, wherein the height of the midpoint of the vegetation point cloud is equivalent to the vegetation height.
2) Single-wood point cloud segmentation: dividing the vegetation point cloud into a plurality of layers according to the height distribution condition of the normalized vegetation point cloud, setting a neighborhood detection window size in each layer, carrying out height local maximum value detection, carrying out three-dimensional space point K-Means clustering on the vegetation point cloud of each layer by taking a local maximum value point as an initial clustering center, judging whether the horizontal distance between clustering centers in each clustering point cloud of adjacent upper and lower 2 layers is smaller than a set clustering center distance threshold value or not from the uppermost layer after the point cloud clustering of each layer is completed, merging the corresponding clustering point clouds in the upper and lower 2 layers if the horizontal distance is smaller than the threshold value until all the layers are completely compared and merged, and finally obtaining the single-tree segmentation point cloud;
3) parameter extraction: counting the highest point height of each single-tree point cloud to obtain single-tree segmentation point clouds, taking the highest point height as the extracted tree height, taking the horizontal coordinate of the point as a single-tree positioning coordinate, calculating the convex hull area of each single-tree segmentation point cloud projected onto the horizontal plane, taking the area as the nearly circular projection area of the tree crown, and calculating the diameter of a circle as the average crown width of the single tree;
2.2 forest stand parameter extraction
1) The plant and tree density is as follows: taking a square sample plot as an investigation area, counting forest plants and trees of the sample plot according to the sample plot single tree division result, and calculating the density value of the plants and trees of the sample plot by using a formula (1);
N＝n/S (1)
wherein N is the plant-tree density (plant/hm)2) (ii) a n is the number of tree plants in the sample plot; s is the area of the sample plot (hm)2)；
2) Average height of forest stand: the forest stand average height extraction method is characterized in that the correlation between the height of a quartile position on sample plot normalized vegetation point cloud data and the height of an actually measured tree is high;
firstly, selecting a certain amount of sample plot survey data as training samples, and calculating the height of the quartile on the sample plot normalized vegetation point cloud, wherein the specific calculation is as follows: ordering the normalized vegetation point clouds in the sample plot range according to the heights, then calculating the heights of upper quartiles of the total height, namely the heights of the upper quartiles of the normalized vegetation point clouds of the sample plot, establishing a linear regression equation between the heights of the upper quartiles of the normalized vegetation point clouds of the sample plot and the average height of the actually measured forest stand, and then carrying out precision verification by using reserved test samples; finally, realizing forest stand average height extraction of the sample plot to be detected according to the sample plot point cloud and a regression equation;
the actual measurement forest stand average tree height is calculated by adopting a cross-section area weighting calculation method, and the calculation formula is as follows:
wherein H is the average stand height HiIs the height of the ith tree, giThe chest height cross-sectional area of the ith tree is shown, and k is the number of forest stand plants.
A square forest sample plot in a certain region of Yangzhou city is selected as a survey area for monitoring, any 3 groups of monitoring points are selected for real-time monitoring, and the monitoring data are as follows:
the inversion method can effectively reduce errors, is simple and feasible, is particularly suitable for inverting forest parameters, is closer to the actual height for the inversion result of the tree height, and can be more beneficial to the actual application of remote sensing images.
It should be noted that the above-mentioned embodiments illustrate rather than limit the technical solutions of the present invention, and that equivalent substitutions and other modifications made by persons skilled in the art according to the prior art are included in the scope of the claims of the present invention as long as they do not exceed the spirit and scope of the technical solutions of the present invention.
Claims (2)
1. A forest surface feature imaging inversion method based on an interferometric SAR is characterized by comprising the following steps: the method comprises the following steps:
step 1, acquiring a group of SAR image data of a target to be detected, extracting topographic information by using a point cloud filtering algorithm, and removing non-ground points from an original point cloud;
an asymptotic irregular triangulation network encryption filtering algorithm is adopted, and the method specifically comprises the following steps:
1) carrying out median filtering denoising on the original point cloud data to remove noise point cloud;
2) meshing point clouds, namely setting the size of a grid, dividing the denoised point clouds into regular grids according to x and y coordinates of space points in the point clouds, and taking the lowest elevation point in each regular grid as an initial ground seed point;
3) taking the initial ground seed points as seed points for constructing an initial irregular triangulation network, introducing slope threshold value judgment in the process of constructing the initial irregular triangulation network, namely firstly performing slope calculation on adjacent network construction nodes when constructing the triangulation network, adding the network construction nodes if the slope threshold value is smaller than the slope threshold value, otherwise marking the nodes as non-ground points, and not participating in calculation in the later steps to finally obtain the initial irregular triangulation network;
4) on the basis of the initial irregular triangulation network, judging the gradient and the altitude difference threshold value of points which are not scratched into the ground point cloud, if the conditions are met, scratching into the ground points, and carrying out iterative encryption on the ground irregular triangulation network;
5) repeating the step 4), stopping iteration when no new ground point is added into the triangulation network, and obtaining triangulation network point cloud meeting the conditions, namely the final ground point cloud;
step 2, single tree parameter extraction and forest stand parameter extraction, wherein the single tree parameters mainly comprise single tree height, crown width and single tree position, and the forest stand parameters mainly comprise plant tree density and forest stand average height;
2.1, extracting parameters of the single wood:
obtaining ground point clouds and non-ground point clouds contained in forest region point cloud data on the basis of the point cloud filtering processing in the step 1, wherein the non-ground point clouds in the forest region are basically equivalent to vegetation point clouds, performing elevation normalization processing on the original vegetation point clouds, determining the elevation of each space point in the output point clouds to be the absolute height of the point relative to the ground, then performing single-tree point cloud segmentation by using the normalized point clouds, and finally extracting the tree height and the crown width according to the single-tree point cloud segmentation result;
the method comprises the following steps of (1) carrying out single-tree point cloud segmentation by adopting a K-Means clustering point cloud segmentation algorithm based on normalized vegetation point cloud height layering:
1) point cloud normalization: dividing vegetation point cloud data obtained by point cloud filtering into different grids according to the generated DEM image resolution size by taking a horizontal X, Y coordinate as a basis, and subtracting the value of the corresponding DEM grid from the elevation of each vegetation point falling in the grids to obtain normalized vegetation point cloud, wherein the height of the midpoint of the vegetation point cloud is equivalent to the vegetation height;
2) single-wood point cloud segmentation: dividing the vegetation point cloud into a plurality of layers according to the height distribution condition of the normalized vegetation point cloud, setting a neighborhood detection window size in each layer, carrying out height local maximum value detection, carrying out three-dimensional space point K-Means clustering on the vegetation point cloud of each layer by taking a local maximum value point as an initial clustering center, judging whether the horizontal distance between clustering centers in each clustering point cloud of the upper and lower 2 layers adjacent to each other is smaller than a set clustering center distance threshold value or not from the uppermost layer after the point cloud clustering of each layer is completed, merging the corresponding clustering point clouds in the upper and lower 2 layers if the horizontal distance is smaller than the threshold value until all the layers are completely compared and merged, and finally obtaining the single-wood segmentation point cloud;
3) parameter extraction: counting the highest point height of each single-tree point cloud of the obtained single-tree segmentation point clouds to be used as the extracted tree height, using the horizontal coordinate of the point as a single-tree positioning coordinate, calculating the convex hull area of each single-tree segmentation point cloud projected on the horizontal plane, using the area as the nearly circular projection area of the tree crown, and calculating the diameter of a circle to be used as the average crown width of the single tree;
2.2 forest stand parameter extraction
1) The plant and tree density is as follows: taking the square sample plot as an investigation area, counting the forest trees in the sample plot according to the single tree division result of the sample plot, and calculating the density value of the trees in the sample plot by the formula (1);
N＝n/S (1)
wherein N is the plant-tree density (plant/hm)2) (ii) a n is the number of tree plants in the sample plot; s is the area of the sample plot (hm)2)；
2) Average height of forest stand: the forest stand average height extraction method is characterized in that the correlation between the height of a quartile position on sample plot normalized vegetation point cloud data and the height of an actually measured tree is high;
firstly, selecting a certain amount of sample plot survey data as training samples, and calculating the height of the quartile on the sample plot normalized vegetation point cloud, wherein the specific calculation is as follows: ordering the normalized vegetation point clouds in the sample plot range according to the heights, then calculating the heights of upper quartiles of the total height, namely the heights of the upper quartiles of the normalized vegetation point clouds of the sample plot, establishing a linear regression equation with the actually measured forest stand average height, and then performing precision verification by using reserved test samples; and finally, realizing forest stand average height extraction of the sample plot to be detected according to the sample plot point cloud and the regression equation.
2. The forest terrain imaging inversion method based on interferometric SAR radar as claimed in claim 1, characterized in that: the actually measured forest stand average tree height is calculated by adopting a cross-section area weighting calculation method, and the calculation formula is as follows:
wherein H is the average stand height HiIs the height of the ith tree, giThe chest height cross-sectional area of the ith tree is shown, and k is the number of forest stand plants.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110057861.3A CN113009481A (en) | 2021-01-15 | 2021-01-15 | Forest surface feature imaging inversion method based on interferometric SAR radar |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110057861.3A CN113009481A (en) | 2021-01-15 | 2021-01-15 | Forest surface feature imaging inversion method based on interferometric SAR radar |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113009481A true CN113009481A (en) | 2021-06-22 |
Family
ID=76384219
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110057861.3A Pending CN113009481A (en) | 2021-01-15 | 2021-01-15 | Forest surface feature imaging inversion method based on interferometric SAR radar |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113009481A (en) |
Cited By (5)
* Cited by examiner, † Cited by third party| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113963160A (en) * | 2021-10-21 | 2022-01-21 | 国网天津市电力公司电力科学研究院 | A fully automatic segmentation method of point cloud based on the spatial position of point cloud |
| CN114266987A (en) * | 2022-03-03 | 2022-04-01 | 水利部长江勘测技术研究所 | Intelligent identification method for high slope dangerous rock mass of unmanned aerial vehicle |
| CN114548277A (en) * | 2022-02-22 | 2022-05-27 | 电子科技大学 | Method and system for fitting ground points and extracting crop height based on point cloud data |
| CN115063677A (en) * | 2022-06-10 | 2022-09-16 | 安徽农业大学 | Wheat field lodging degree identification method and device based on point cloud information |
| CN117765401A (en) * | 2024-01-11 | 2024-03-26 | 航天信德智图（北京）科技有限公司 | Forest parameter extraction method, device, equipment and medium based on multi-source remote sensing |
Citations (16)
* Cited by examiner, † Cited by third party| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102401898A (en) * | 2011-08-25 | 2012-04-04 | 北京理工大学 | Quantified simulation method for forest remote sensing data of synthetic aperture radar |
| CN103996175A (en) * | 2014-05-13 | 2014-08-20 | 西安电子科技大学 | Forest or urban area high-resolution interference phase filtering method |
| CN104867180A (en) * | 2015-05-28 | 2015-08-26 | 南京林业大学 | UAV and LiDAR integrated forest stand characteristic inversion method |
| CN105005047A (en) * | 2015-07-17 | 2015-10-28 | 武汉大学 | Forest complex terrain correction and forest height inversion methods and systems with backscattering optimization |
| CN107479065A (en) * | 2017-07-14 | 2017-12-15 | 中南林业科技大学 | A kind of three-dimensional structure of forest gap method for measurement based on laser radar |
| CA3028602A1 (en) * | 2016-08-01 | 2018-02-08 | Mitsubishi Electric Corporation | Synthetic-aperture radar device |
| CN109061601A (en) * | 2018-08-09 | 2018-12-21 | 南京林业大学 | A method of based on unmanned plane laser radar inverting artificial forest forest structural variable |
| CN109164459A (en) * | 2018-08-01 | 2019-01-08 | 南京林业大学 | A kind of method that combination laser radar and high-spectral data classify to forest species |
| CN109711446A (en) * | 2018-12-18 | 2019-05-03 | 中国科学院深圳先进技术研究院 | A method and device for classifying ground objects based on multispectral images and SAR images |
| CN109816779A (en) * | 2019-01-30 | 2019-05-28 | 北京林业大学 | A method for obtaining single-tree parameters by reconstructing a plantation forest model using a smartphone |
| CN109946714A (en) * | 2019-04-03 | 2019-06-28 | 海南省林业科学研究所 | A kind of method for building up of the forest biomass model based on LiDAR and ALOS PALSAR multivariate data |
| CN110109111A (en) * | 2019-04-28 | 2019-08-09 | 西安电子科技大学 | Polarimetric SAR interferometry sparse vegetation height inversion method |
| CN110223314A (en) * | 2019-06-06 | 2019-09-10 | 电子科技大学 | A kind of single wooden dividing method based on the distribution of tree crown three-dimensional point cloud |
| CN110570428A (en) * | 2019-08-09 | 2019-12-13 | 浙江合信地理信息技术有限公司 | method and system for segmenting roof surface patch of building from large-scale image dense matching point cloud |
| CN111091079A (en) * | 2019-12-04 | 2020-05-01 | 生态环境部南京环境科学研究所 | TLS-based method for measuring dominant single plant structural parameters of vegetation in alpine and fragile regions |
| CN111553245A (en) * | 2020-04-24 | 2020-08-18 | 中国电建集团成都勘测设计研究院有限公司 | Vegetation classification method based on machine learning algorithm and multi-source remote sensing data fusion |
-
2021
- 2021-01-15 CN CN202110057861.3A patent/CN113009481A/en active Pending
Patent Citations (16)
* Cited by examiner, † Cited by third party| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102401898A (en) * | 2011-08-25 | 2012-04-04 | 北京理工大学 | Quantified simulation method for forest remote sensing data of synthetic aperture radar |
| CN103996175A (en) * | 2014-05-13 | 2014-08-20 | 西安电子科技大学 | Forest or urban area high-resolution interference phase filtering method |
| CN104867180A (en) * | 2015-05-28 | 2015-08-26 | 南京林业大学 | UAV and LiDAR integrated forest stand characteristic inversion method |
| CN105005047A (en) * | 2015-07-17 | 2015-10-28 | 武汉大学 | Forest complex terrain correction and forest height inversion methods and systems with backscattering optimization |
| CA3028602A1 (en) * | 2016-08-01 | 2018-02-08 | Mitsubishi Electric Corporation | Synthetic-aperture radar device |
| CN107479065A (en) * | 2017-07-14 | 2017-12-15 | 中南林业科技大学 | A kind of three-dimensional structure of forest gap method for measurement based on laser radar |
| CN109164459A (en) * | 2018-08-01 | 2019-01-08 | 南京林业大学 | A kind of method that combination laser radar and high-spectral data classify to forest species |
| CN109061601A (en) * | 2018-08-09 | 2018-12-21 | 南京林业大学 | A method of based on unmanned plane laser radar inverting artificial forest forest structural variable |
| CN109711446A (en) * | 2018-12-18 | 2019-05-03 | 中国科学院深圳先进技术研究院 | A method and device for classifying ground objects based on multispectral images and SAR images |
| CN109816779A (en) * | 2019-01-30 | 2019-05-28 | 北京林业大学 | A method for obtaining single-tree parameters by reconstructing a plantation forest model using a smartphone |
| CN109946714A (en) * | 2019-04-03 | 2019-06-28 | 海南省林业科学研究所 | A kind of method for building up of the forest biomass model based on LiDAR and ALOS PALSAR multivariate data |
| CN110109111A (en) * | 2019-04-28 | 2019-08-09 | 西安电子科技大学 | Polarimetric SAR interferometry sparse vegetation height inversion method |
| CN110223314A (en) * | 2019-06-06 | 2019-09-10 | 电子科技大学 | A kind of single wooden dividing method based on the distribution of tree crown three-dimensional point cloud |
| CN110570428A (en) * | 2019-08-09 | 2019-12-13 | 浙江合信地理信息技术有限公司 | method and system for segmenting roof surface patch of building from large-scale image dense matching point cloud |
| CN111091079A (en) * | 2019-12-04 | 2020-05-01 | 生态环境部南京环境科学研究所 | TLS-based method for measuring dominant single plant structural parameters of vegetation in alpine and fragile regions |
| CN111553245A (en) * | 2020-04-24 | 2020-08-18 | 中国电建集团成都勘测设计研究院有限公司 | Vegetation classification method based on machine learning algorithm and multi-source remote sensing data fusion |
Non-Patent Citations (2)
* Cited by examiner, † Cited by third party| Title |
|---|
| 于大洋 等: "基于干涉极化SAR数据的森林树高反演", 《清华大学学报(自然科学版)》 * |
| 邵为真 等: "基于不规则三角网的渐进加密滤波算法研究", 《北京测绘》 * |
Cited By (7)
* Cited by examiner, † Cited by third party| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113963160A (en) * | 2021-10-21 | 2022-01-21 | 国网天津市电力公司电力科学研究院 | A fully automatic segmentation method of point cloud based on the spatial position of point cloud |
| CN114548277A (en) * | 2022-02-22 | 2022-05-27 | 电子科技大学 | Method and system for fitting ground points and extracting crop height based on point cloud data |
| CN114548277B (en) * | 2022-02-22 | 2023-09-08 | 电子科技大学 | Method and system for ground point fitting and crop height extraction based on point cloud data |
| CN114266987A (en) * | 2022-03-03 | 2022-04-01 | 水利部长江勘测技术研究所 | Intelligent identification method for high slope dangerous rock mass of unmanned aerial vehicle |
| CN115063677A (en) * | 2022-06-10 | 2022-09-16 | 安徽农业大学 | Wheat field lodging degree identification method and device based on point cloud information |
| CN115063677B (en) * | 2022-06-10 | 2023-10-10 | 安徽农业大学 | A method and device for identifying the lodging degree of wheat fields based on point cloud information |
| CN117765401A (en) * | 2024-01-11 | 2024-03-26 | 航天信德智图（北京）科技有限公司 | Forest parameter extraction method, device, equipment and medium based on multi-source remote sensing |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113009481A (en) | 2021-06-22 | Forest surface feature imaging inversion method based on interferometric SAR radar |
| CN112381861B (en) | 2024-04-16 | Forest land point cloud data registration and segmentation method based on foundation laser radar |
| Jin et al. | 2018 | Stem–leaf segmentation and phenotypic trait extraction of individual maize using terrestrial LiDAR data |
| Hamraz et al. | 2016 | A robust approach for tree segmentation in deciduous forests using small-footprint airborne LiDAR data |
| CN111898688B (en) | 2023-12-05 | Airborne LiDAR data tree classification method based on three-dimensional deep learning |
| Van Leeuwen et al. | 2010 | Retrieval of forest structural parameters using LiDAR remote sensing |
| Hladik et al. | 2013 | Salt marsh elevation and habitat mapping using hyperspectral and LIDAR data |
| Zhou et al. | 2020 | Individual tree parameters estimation for plantation forests based on UAV oblique photography |
| CN109446986B (en) | 2021-09-24 | An efficient feature extraction and tree species identification method for tree laser point cloud |
| Chen et al. | 2018 | Estimation of forest leaf area index using terrestrial laser scanning data and path length distribution model in open-canopy forests |
| CN105389538A (en) | 2016-03-09 | Method for estimating forest leaf-area index based on point cloud hemisphere slice |
| CN109446983A (en) | 2019-03-08 | A method for estimating the logging volume of coniferous forests based on two-phase UAV images |
| CN107479065A (en) | 2017-12-15 | A kind of three-dimensional structure of forest gap method for measurement based on laser radar |
| CN111091079A (en) | 2020-05-01 | TLS-based method for measuring dominant single plant structural parameters of vegetation in alpine and fragile regions |
| CN112150479A (en) | 2020-12-29 | Single tree segmentation and tree height and crown width extraction method based on Gaussian clustering |
| CN114299318A (en) | 2022-04-08 | Method and system for rapid point cloud data processing and target image matching |
| CN117455963A (en) | 2024-01-26 | Natural forest region foundation airborne laser point cloud registration method |
| CN117197677A (en) | 2023-12-08 | Tropical rain forest arbor-shrub separation method based on laser radar point cloud data |
| CN117765006A (en) | 2024-03-26 | Multi-level dense crown segmentation method based on unmanned aerial vehicle image and laser point cloud |
| Liang et al. | 2022 | Mapping urban impervious surface with an unsupervised approach using interferometric coherence of SAR images |
| CN109946714A (en) | 2019-06-28 | A kind of method for building up of the forest biomass model based on LiDAR and ALOS PALSAR multivariate data |
| Chai et al. | 2023 | A novel solution for extracting individual tree crown parameters in high-density plantation considering inter-tree growth competition using terrestrial close-range scanning and photogrammetry technology |
| Yip et al. | 2024 | Community-based plant diversity monitoring of a dense-canopy and species-rich tropical forest using airborne LiDAR data |
| Zhu et al. | 2021 | Research on deep learning individual tree segmentation method coupling RetinaNet and point cloud clustering |
| CN108132096B (en) | 2020-07-07 | A method for monitoring solar radiation of forest windows based on lidar |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| 2021-06-22 | PB01 | Publication | |
| 2021-06-22 | PB01 | Publication | |
| 2021-07-09 | SE01 | Entry into force of request for substantive examination | |
| 2021-07-09 | SE01 | Entry into force of request for substantive examination | |
| 2023-03-17 | AD01 | Patent right deemed abandoned |
Effective date of abandoning: 20230317 |
| 2023-03-17 | AD01 | Patent right deemed abandoned |