CN112396615A - Individual stumpage segmentation method and device based on adjustable hemispherical canopy camera - Google Patents
Individual stumpage segmentation method and device based on adjustable hemispherical canopy camera Download PDFInfo
- Publication number
- CN112396615A CN112396615A CN202011364465.7A CN202011364465A CN112396615A CN 112396615 A CN112396615 A CN 112396615A CN 202011364465 A CN202011364465 A CN 202011364465A CN 112396615 A CN112396615 A CN 112396615A
- Authority
- CN
- China
- Prior art keywords
- segmentation
- tree
- crown
- camera
- canopy
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 230000011218 segmentation Effects 0.000 title claims abstract description 79
- 238000000034 method Methods 0.000 title claims abstract description 32
- 238000011160 research Methods 0.000 claims abstract description 11
- 230000003595 spectral effect Effects 0.000 claims description 13
- 238000000605 extraction Methods 0.000 claims description 11
- 241000196324 Embryophyta Species 0.000 claims description 6
- 238000003709 image segmentation Methods 0.000 claims description 5
- 238000012360 testing method Methods 0.000 claims description 4
- 230000000007 visual effect Effects 0.000 claims description 4
- 239000002023 wood Substances 0.000 claims description 4
- 238000012549 training Methods 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 abstract description 3
- 238000012986 modification Methods 0.000 description 5
- 230000004048 modification Effects 0.000 description 5
- 230000008901 benefit Effects 0.000 description 3
- 238000013461 design Methods 0.000 description 3
- 238000011161 development Methods 0.000 description 3
- 230000018109 developmental process Effects 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 238000004519 manufacturing process Methods 0.000 description 3
- 238000005259 measurement Methods 0.000 description 3
- 238000005520 cutting process Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 230000000877 morphologic effect Effects 0.000 description 1
- 238000012805 post-processing Methods 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 238000012552 review Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 241000894007 species Species 0.000 description 1
- 238000001228 spectrum Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2413—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on distances to training or reference patterns
- G06F18/24147—Distances to closest patterns, e.g. nearest neighbour classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/194—Segmentation; Edge detection involving foreground-background segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10024—Color image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20112—Image segmentation details
- G06T2207/20152—Watershed segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30181—Earth observation
- G06T2207/30188—Vegetation; Agriculture
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Multimedia (AREA)
- Image Processing (AREA)
Abstract
The invention relates to the technical field of hyperspectral single tree segmentation, in particular to a single standing tree segmentation method and a single standing tree segmentation device based on an adjustable hemispherical canopy camera. The method for segmenting the single standing tree based on the adjustable hemispherical canopy camera comprises the following steps: s1, selecting an optimal segmentation parameter combination, segmenting the DOM of the research area according to the optimal segmentation parameter combination, and preparing for extracting the single-tree crown; and S2, extracting the information of the single tree crown, classifying the image after the image is segmented, and finally outputting the vector file of the single tree crown according to the classification. The method for segmenting the single standing tree based on the adjustable hemispherical canopy camera can be simultaneously suitable for high-density tree areas and low-density tree areas, the application range of the method is widened, the improved graph node similarity calculation method enables the branch tip to have a better segmentation result, and the detail degree of the segmentation result is guaranteed.
Description
Technical Field
The invention relates to the technical field of hyperspectral single tree segmentation, in particular to a single standing tree segmentation method and a single standing tree segmentation device based on an adjustable hemispherical canopy camera.
Background
In China, forest Canopy identification is usually performed based on an airborne radar, Chinese patent CN107368813A discloses a forest Canopy width identification method applied to an airborne near-earth hyperspectral image, a large amount of actual measured single tree data is utilized in single tree segmentation research based on airborne radar data, a correlation equation of tree Height and Canopy width is calculated after classification is performed according to Height, the size of a filter window in the single tree extraction process is determined according to the equation, then a Canopy Height Model (Canopy Height Model, CHM) is established by combining the airborne radar data through an interpolation method, each pixel value in the CHM is brought into the correlation equation of tree Height and Canopy width, the size of the filter window corresponding to the pixel is calculated, meanwhile, the maximum value in the window is judged, and if the maximum value is taken as the highest point of the tree, the tree is identified. However, the prior art has the problems of inaccurate measurement, low single-plant standing tree segmentation precision and single function of measuring equipment.
Disclosure of Invention
In order to solve the problems in the prior art, the invention provides a method and a device for cutting single standing tree based on an adjustable hemispherical canopy camera, which are accurate in measurement, high in cutting accuracy of the single standing tree and more practical in measurement equipment.
In order to solve the technical problems, the invention provides the following technical scheme:
a single standing tree segmentation method based on an adjustable hemispherical canopy camera comprises the following steps:
s1, selecting an optimal segmentation parameter combination, segmenting the DOM of the research area according to the optimal segmentation parameter combination, and preparing for extracting the single-tree crown;
and S2, extracting the information of the single tree crown, classifying the image after the image is segmented, and finally outputting the vector file of the single tree crown according to the classification.
Further, the image segmentation comprises the following steps:
forming the image object into a polygon, firstly establishing a plurality of categories, and respectively selecting a plurality of samples for training for each category by using a closest distance method;
then merging the vegetation polygons based on classes, and merging the divided images of all the vegetation into vegetation information polygons again to prepare for further segmentation;
according to the difference of spectral characteristics and shape attributes of arbors and other vegetation, member function setting is carried out, shrub and weed information among crowns is removed, and image segmentation is carried out to extract arbors crown polygons;
and outputting the vector format file of the divided single-tree crown as test crown data.
Further, several of the categories include roads, buildings, ponds, and vegetation.
Further, the membership function setting is performed by using the attribute of the contrast of the adjacent pixels.
Further, the single-tree crown information extraction method comprises the following steps:
correcting the boundary of the tree crown, segmenting to obtain a single tree crown, setting NDVI parameters, and setting non-tree crown objects which meet the rule that NDVI is less than 0.35& chm and less than 800;
setting a spectral factor to be 0.1, a compactness factor to be 0.9 and a segmentation scale to be 60-80;
the weights of red, green, blue and CHM are respectively set as 1, 1 and 0.1, and the segmentation scale, the spectral factor and the compactness factor are respectively 76, 0.1 and 0.9 which are the same as the optimal segmentation.
Further, the single-tree crown information extraction also comprises the extraction of watershed transformation, and the method comprises the following steps:
adjusting the local minimum value, and setting the radius of the circular structural element to be 2 so as to reduce the number of the local minimum values;
marking foreground and background objects, wherein the foreground object is a crown area;
and carrying out watershed segmentation on the foreground object.
Further, the watershed segmentation of the foreground object comprises the following steps:
selecting a sample area, and visually identifying and drawing a crown boundary vector;
determining the identification number of the sample region crown;
superimposing the manual segmentation vector on the results of the multi-scale segmentation and the watershed segmentation;
several visual identification crowns were segmented.
The invention also provides the following technical scheme:
the utility model provides a single trunk founds wood segmenting device based on adjustable hemisphere canopy camera, includes unmanned aerial vehicle (100) and sets up acquisition unit (200) on unmanned aerial vehicle (100).
Further, the unmanned aerial vehicle (100) comprises a fuselage (101), a cantilever (102) connected with the fuselage (101), and a wing connected to the cantilever (102); the acquisition unit (200) comprises a connecting piece (300) connected with the machine body (101) and a camera assembly (400) connected with the connecting piece (300).
Further, the camera assembly (400) is a video camera.
Compared with the prior art, the invention has the beneficial effects that:
the airborne small-spot laser radar point cloud data and the large-scale aerial photo synchronously acquired by the airborne small-spot laser radar point cloud data are taken as data sources, theoretically, high-precision single tree parameter information can be extracted, and the results prove that the high-precision single tree parameter information can be extracted by combining LiDAR data with the large-scale aerial photo, and in single tree level segmentation, the trunk detection strategy of histogram morphological characteristics and ground connectivity is combined, so that the method can be simultaneously suitable for high-density tree areas and low-density tree areas, and the application range of the method is widened; in crown level segmentation, the improved graph node similarity calculation method enables branch tips to have better segmentation results, and ensures the detail degree of the segmentation results.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the description of the embodiments will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without inventive exercise. Wherein:
FIG. 1 is a schematic diagram of a single tree crown extraction process in a single standing tree segmentation method based on an adjustable hemispherical canopy camera according to the present invention;
fig. 2 is a schematic structural diagram of a single standing tree splitting device based on an adjustable hemispherical canopy camera according to the present invention.
Detailed Description
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in detail below.
In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention, but the present invention may be practiced in other ways than those specifically described and will be readily apparent to those of ordinary skill in the art without departing from the spirit of the present invention, and therefore the present invention is not limited to the specific embodiments disclosed below.
Next, the present invention will be described in detail with reference to the drawings, wherein the cross-sectional views illustrating the structure of the device are not enlarged partially according to the general scale for convenience of illustration when describing the embodiments of the present invention, and the drawings are only examples, which should not limit the scope of the present invention. In addition, the three-dimensional dimensions of length, width and depth should be included in the actual fabrication.
Further, reference herein to "one embodiment" or "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one implementation of the invention. The appearances of the phrase "in one embodiment" in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments.
The embodiment of the invention comprises the following steps:
example 1:
the embodiment provides a single standing tree segmentation method based on an adjustable hemispherical canopy camera, along with the continuous development of remote sensing technology, the spatial resolution of remote sensing images is continuously increased, the spatial resolution of the current aerial remote sensing number images is broken through the decimeter level, more and more abundant ground detail features can be seen from the images, narrow and small targets which are difficult to distinguish on low-resolution images are also distinguished, and the single standing tree distinguishing in large-scale aerial photos can be realized. The single standing tree segmentation is to extract the crown information of the trees as completely as possible on a high-resolution remote sensing image. The difficulty is that the crown of the forest tree and the surrounding low vegetation are extracted simultaneously, so that the area of the crown is increased, or the crown is overlapped, so that the divided crown is smaller than the actual crown. In order to accurately segment the single standing tree, an object-oriented segmentation method is used for research, an optimal segmentation parameter combination is obtained through a multi-scale segmentation test, the single standing tree theme is segmented from a large theme of a protected area, and data preparation is made for the next estimation of the single standing tree parameters.
The method for segmenting the single standing tree based on the adjustable hemispherical canopy camera comprises the following steps:
s1, selecting an optimal segmentation parameter combination, segmenting the DOM of the research area according to the optimal segmentation parameter combination, and preparing for extracting the single-tree crown;
and S2, extracting the information of the single tree crown, classifying the image after the image is segmented, and finally outputting the vector file of the single tree crown according to the classification.
Through experiments, the DOM of the research area is segmented under different segmentation parameter combinations, the parameter combination with ideal segmentation effect is selected, and the optimal segmentation parameter combination suitable for the research area is finally selected, as shown in Table 1:
TABLE 1
| Purpose of division | Spectral factor | Form factor | Smoothness of the surface | Compactness degree |
| Differentiating vegetation from non-vegetation | 0.9 | 0.1 | 0.2 | 0.8 |
| Distinguishing single wood | 0.6 | 0.4 | 0.8 | 0.2 |
And segmenting the DOM of the research area according to the optimal segmentation parameter combination, and preparing for extracting the single-tree crown.
In this embodiment, the step of completing the segmentation of the image comprises the following steps:
forming the image object into a polygon, firstly establishing a plurality of categories, wherein the categories comprise a road category, a building category, a pond category and a vegetation category, and respectively selecting a plurality of samples for training for each category by using a closest distance method;
then merging the vegetation polygons based on classes, and merging the divided images of all the vegetation into vegetation information polygons again to prepare for further segmentation;
the main purpose of the segmentation is to distinguish vegetation from non-vegetation, and the spectral characteristics of the vegetation and other three land species are obviously different, so that the weight of the selected spectral factor is great, and the weight of the selected form factor is small. The non-vegetation type is mostly rectangular, so the weight of compactness is big, and the weight of smoothness is little, and the image object polygon of dividing relatively accords with actual conditions like this. Because the condition of the forest area is complex, the canopy density is greatly different between the natural forest and the artificial forest as well as between the natural forest and the artificial forest, and the forest age, the forest gap can be obviously seen on a 0.2m aviation lens. However, in the process of dividing the forest and other ground features, shrubs, weeds and the like around the forest are also included in the vegetation information polygon, and the second division aims to obtain the polygon of the crown of a single arbor, and the shrubs and the weed information need to be distinguished;
the object of image segmentation again is a vegetation polygon, member functions are set by using the attribute of adjacent pixel contrast according to the difference of the spectral characteristics and the shape attributes of trees and other vegetation, shrubs and weed information among crowns are removed, and the tree crown polygon is extracted by image segmentation; in this division, the spectral characteristics of the arbor and shrub weeds are not significantly different, but the crown of each arbor appears as a relatively smooth polygon on the image, so the weight of the shape factor is increased, and the weight of the smoothness is 0.8. Extracting the cut crown information into edges, and performing post-processing on the crown edges, namely setting a polygon area threshold value by an area screening method, extracting fine and broken small polygons which are wrongly extracted into the edges, and merging the fine and broken small polygons into the crown by category merging;
and finally, manually editing the missed or mistaken tree crown polygon, so that the single-tree scale segmentation of the research area is completed.
And then outputting the vector format file of the divided single-tree crown as test crown data.
In this embodiment, a single-tree crown extraction process is shown in fig. 2, a first step of multi-stage segmentation is to classify a crown part and a ground, a small segmentation scale is selected, an object with CHM >100cm is defined as a forest, and other ground feature points are mistakenly classified as forests.
The second step of extracting the single tree crown information is as follows: and (3) correcting the boundary of the tree crown, segmenting to obtain the single-tree crown, setting NDVI parameters, and setting non-tree crown objects which meet the rule that NDVI is less than 0.35& chm and less than 800.
And thirdly, the single-tree crown is obtained through segmentation, the spectral factor is set to be 0.1, the compactness factor is set to be 0.9, the segmentation scale is set to be 60-80, the whole is observed, the segmentation scale is found to be 60-80, the effect is good, and the crown segmentation effect is best when the segmentation scale is finally determined to be 76.
And adding a CHM layer during segmentation, wherein the standard deviation of the CHM layer is more than 10 times of that of a red-green-blue wave band, the weights of red, green, blue and CHM are respectively set to be 1, 1 and 0.1, and the segmentation scale, the spectral factor and the compactness factor are respectively 76, 0.1 and 0.9 which are the same as the optimal segmentation above for segmentation.
In this embodiment, the single-tree crown information extraction further includes extraction of watershed transform, including the following steps:
adjusting the local minimum value, and setting the radius of the circular structural element to be 2 so as to reduce the number of the local minimum values;
marking foreground and background objects, wherein the foreground object is a crown area;
and carrying out watershed segmentation on the foreground object.
In this embodiment, performing watershed segmentation on the foreground object includes the following steps:
selecting a sample area, and visually identifying and drawing a crown boundary vector;
determining the identification number of the sample region crown;
superimposing the manual segmentation vector on the results of the multi-scale segmentation and the watershed segmentation;
counting the divided situations of a plurality of visual identification crowns, and finally finding out the part with the most multi-scale division as the crown is divided into three objects.
The reason is as follows: firstly, because the area of the crown width of the trees in the area has great difference, and secondly, because most of the trees in the area are coniferous forests, the trees are very easily affected by the sunlight and are often separated from the tops of the trees.
The most existing watershed segmentation is that the multi-crown is divided into the same object.
The reason is as follows: firstly, the generated CHM has low precision due to the low density of the laser point cloud; secondly, because the crowns in the region are covered and overlapped, multiple crowns are divided into the same object, and the division result is less than the number of actual crowns.
In this embodiment, the polygon of the single-tree crown divided by the DOM is subjected to edge smoothing processing and then output in a vector format, and the diameter of a circle having the same area as the polygon is used as the amplitude of the crown. In actual ground sample survey, the center coordinates of a sample are located by a differential GPS, the positional relationship between a sample and the center point of the sample is recorded by a tape and a compass, and the position coordinates of each sample are calculated in data processing. And (3) superposing the sample plot center coordinate and sample tree coordinate generation vector file with the orthoimage of the research area, wherein because the forest area has large positioning deviation by using a GPS, visual comparison is also carried out, each tree is accurately positioned, the polygonal area of the crown is output, and the diameter of the tree is reversely deduced through a circle area calculation formula to be used as the crown width of the single tree to be researched.
Example 2:
as shown in fig. 2, a second embodiment of the present invention is based on the previous embodiment, and is different from the previous embodiment in that:
the invention also provides the following technical scheme:
the utility model provides a single trunk founds wood segmenting device based on adjustable hemisphere canopy camera, includes unmanned aerial vehicle (100) and sets up acquisition unit (200) on unmanned aerial vehicle (100).
As shown in fig. 2, the drone (100) comprises a fuselage (101), a cantilever (102) connected to the fuselage (101), and a wing connected to the cantilever (102); the acquisition unit (200) comprises a connecting piece (300) connected with the machine body (101) and a camera assembly (400) connected with the connecting piece (300).
In this embodiment, the camera assembly (400) is a video camera.
In specific implementation, the invention carries a spectrum/multispectral image space-three positioning technology, and the hyperspectral/multispectral remote sensing image space-three stereo positioning, and has the key that the function model-based front intersection is realized, for the left image and the right image of a stereo pair, after respective function models are respectively established, the space coordinates of corresponding ground points are calculated according to the image coordinates of image points with the same name, and the problem of calculating the coordinates of three-dimensional object space points based on the function models is solved, wherein the result obtained by encrypting control points can obviously improve the precision of geometric correction of the hyperspectral/multispectral remote sensing stereo image.
It is important to note that the construction and arrangement of the present application as shown in the various exemplary embodiments is illustrative only. Although only a few embodiments have been described in detail in this disclosure, those skilled in the art who review this disclosure will readily appreciate that many modifications are possible (e.g., variations in sizes, dimensions, structures, shapes and proportions of the various elements, values of parameters (e.g., temperatures, pressures, etc.), mounting arrangements, use of materials, colors, orientations, etc.) without materially departing from the novel teachings and advantages of the subject matter recited in this application. For example, elements shown as integrally formed may be constructed of multiple parts or elements, the position of elements may be reversed or otherwise varied, and the nature or number of discrete elements or positions may be altered or varied. Accordingly, all such modifications are intended to be included within the scope of this invention. The order or sequence of any process or method steps may be varied or re-sequenced according to alternative embodiments. In the claims, any means-plus-function clause is intended to cover the structures described herein as performing the recited function and not only structural equivalents but also equivalent structures. Other substitutions, modifications, changes and omissions may be made in the design, operating conditions and arrangement of the exemplary embodiments without departing from the scope of the present inventions. Therefore, the present invention is not limited to a particular embodiment, but extends to various modifications that nevertheless fall within the scope of the appended claims.
Moreover, in an effort to provide a concise description of the exemplary embodiments, all features of an actual implementation may not be described (i.e., those unrelated to the presently contemplated best mode of carrying out the invention, or those unrelated to enabling the invention).
It should be appreciated that in the development of any such actual implementation, as in any engineering or design project, numerous implementation-specific decisions may be made. Such a development effort might be complex and time consuming, but would nevertheless be a routine undertaking of design, fabrication, and manufacture for those of ordinary skill having the benefit of this disclosure, without undue experimentation.
It should be noted that the above-mentioned embodiments are only for illustrating the technical solutions of the present invention and not for limiting, and although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that modifications or equivalent substitutions may be made on the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, which should be covered by the claims of the present invention.
Claims (10)
1. A single standing tree segmentation method based on an adjustable hemispherical canopy camera is characterized by comprising the following steps:
s1, selecting an optimal segmentation parameter combination, segmenting the DOM of the research area according to the optimal segmentation parameter combination, and preparing for extracting the single-tree crown;
and S2, extracting the information of the single tree crown, classifying the image after the image is segmented, and finally outputting the vector file of the single tree crown according to the classification.
2. The method for segmenting the single standing tree based on the adjustable hemispherical canopy camera as claimed in claim 1, wherein the step of completing segmentation of the image comprises the following steps:
forming the image object into a polygon, firstly establishing a plurality of categories, and respectively selecting a plurality of samples for training for each category by using a closest distance method;
then merging the vegetation polygons based on classes, and merging the divided images of all the vegetation into vegetation information polygons again to prepare for further segmentation;
according to the difference of spectral characteristics and shape attributes of arbors and other vegetation, member function setting is carried out, shrub and weed information among crowns is removed, and image segmentation is carried out to extract arbors crown polygons;
and outputting the vector format file of the divided single-tree crown as test crown data.
3. The method of claim 2, wherein the categories include roads, buildings, ponds, and vegetation.
4. The adjustable hemispherical canopy camera-based individual standing tree segmentation method of claim 3, wherein membership function settings are made with attributes of adjacent pixel contrast.
5. The method for segmenting the single standing tree based on the adjustable hemispherical canopy camera as claimed in claim 4, wherein the single tree canopy information extraction comprises the following steps:
correcting the boundary of the tree crown, segmenting to obtain a single tree crown, setting NDVI parameters, and setting non-tree crown objects which meet the rule that NDVI is less than 0.35& chm and less than 800;
setting a spectral factor to be 0.1, a compactness factor to be 0.9 and a segmentation scale to be 60-80;
the weights of red, green, blue and CHM are respectively set as 1, 1 and 0.1, and the segmentation scale, the spectral factor and the compactness factor are respectively 76, 0.1 and 0.9 which are the same as the optimal segmentation.
6. The tunable hemispherical canopy camera-based individual standing tree segmentation method of claim 5, wherein the individual tree canopy information extraction further comprises an extraction of watershed transforms, comprising the steps of:
adjusting the local minimum value, and setting the radius of the circular structural element to be 2 so as to reduce the number of the local minimum values;
marking foreground and background objects, wherein the foreground object is a crown area;
and carrying out watershed segmentation on the foreground object.
7. The tunable hemispherical canopy camera-based individual standing tree segmentation method of claim 6, wherein watershed segmentation of foreground objects comprises the steps of:
selecting a sample area, and visually identifying and drawing a crown boundary vector;
determining the identification number of the sample region crown;
superimposing the manual segmentation vector on the results of the multi-scale segmentation and the watershed segmentation;
several visual identification crowns were segmented.
8. The utility model provides a single trunk founds wood segmenting device based on adjustable hemisphere canopy camera which characterized in that, includes unmanned aerial vehicle (100) and sets up acquisition unit (200) on unmanned aerial vehicle (100).
9. The adjustable hemispherical canopy camera-based individual standing tree splitting apparatus of claim 8, wherein the drone (100) comprises a fuselage (101), a boom (102) connected to the fuselage (101), and a wing connected to the boom (102); the acquisition unit (200) comprises a connecting piece (300) connected with the machine body (101) and a camera assembly (400) connected with the connecting piece (300).
10. The adjustable dome canopy camera-based individual stumpage splitting device of claim 9, wherein said camera assembly (400) is a video camera.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011364465.7A CN112396615A (en) | 2020-11-27 | 2020-11-27 | Individual stumpage segmentation method and device based on adjustable hemispherical canopy camera |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011364465.7A CN112396615A (en) | 2020-11-27 | 2020-11-27 | Individual stumpage segmentation method and device based on adjustable hemispherical canopy camera |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112396615A true CN112396615A (en) | 2021-02-23 |
Family
ID=74605434
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011364465.7A Pending CN112396615A (en) | 2020-11-27 | 2020-11-27 | Individual stumpage segmentation method and device based on adjustable hemispherical canopy camera |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112396615A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113592882A (en) * | 2021-08-09 | 2021-11-02 | 中国林业科学研究院资源信息研究所 | Crown extraction method based on multi-source remote sensing of unmanned aerial vehicle |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090316988A1 (en) * | 2008-06-18 | 2009-12-24 | Samsung Electronics Co., Ltd. | System and method for class-specific object segmentation of image data |
| CN104881868A (en) * | 2015-05-14 | 2015-09-02 | 中国科学院遥感与数字地球研究所 | Method for extracting phytocoenosium spatial structure |
| US20170124717A1 (en) * | 2015-10-28 | 2017-05-04 | Intel Corporation | Method and system of background-foreground segmentation for image processing |
| CN106651861A (en) * | 2016-11-15 | 2017-05-10 | 兰州大学 | Method for automatically and optimally selecting remote sensing image segmentation parameters based on regional inconsistency evaluation |
| 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 |
| WO2019198412A1 (en) * | 2018-04-11 | 2019-10-17 | 国立大学法人信州大学 | Forest resources information calculation method and forest resources information calculation device |
| US20200334458A1 (en) * | 2019-04-16 | 2020-10-22 | Precision Silver, LLC | Systems and methods for rating vegetation health and biomass from remotely sensed morphological and radiometric data |
-
2020
- 2020-11-27 CN CN202011364465.7A patent/CN112396615A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090316988A1 (en) * | 2008-06-18 | 2009-12-24 | Samsung Electronics Co., Ltd. | System and method for class-specific object segmentation of image data |
| CN104881868A (en) * | 2015-05-14 | 2015-09-02 | 中国科学院遥感与数字地球研究所 | Method for extracting phytocoenosium spatial structure |
| US20170124717A1 (en) * | 2015-10-28 | 2017-05-04 | Intel Corporation | Method and system of background-foreground segmentation for image processing |
| CN106651861A (en) * | 2016-11-15 | 2017-05-10 | 兰州大学 | Method for automatically and optimally selecting remote sensing image segmentation parameters based on regional inconsistency evaluation |
| 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 |
| WO2019198412A1 (en) * | 2018-04-11 | 2019-10-17 | 国立大学法人信州大学 | Forest resources information calculation method and forest resources information calculation device |
| US20200334458A1 (en) * | 2019-04-16 | 2020-10-22 | Precision Silver, LLC | Systems and methods for rating vegetation health and biomass from remotely sensed morphological and radiometric data |
Non-Patent Citations (2)
| Title |
|---|
| 王平: "基于机载LiDAR数据和航空像片的单木参数提取研究", 中国优秀硕士学位论文全文数据库农业科技辑, no. 1, pages 049 - 159 * |
| 郑鑫等: "基于形态学阈值标记分水岭算法的高分辨率 影像单木树冠提取", 《中南林业调查规划》, vol. 36, no. 4, pages 30 - 35 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113592882A (en) * | 2021-08-09 | 2021-11-02 | 中国林业科学研究院资源信息研究所 | Crown extraction method based on multi-source remote sensing of unmanned aerial vehicle |
| CN113592882B (en) * | 2021-08-09 | 2022-06-03 | 中国林业科学研究院资源信息研究所 | Crown extraction method based on multi-source remote sensing of unmanned aerial vehicle |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Huang et al. | Individual tree crown detection and delineation from very-high-resolution UAV images based on bias field and marker-controlled watershed segmentation algorithms | |
| CN109146889B (en) | Farmland boundary extraction method based on high-resolution remote sensing image | |
| Grigillo et al. | Urban object extraction from digital surface model and digital aerial images | |
| CN113591766B (en) | Multi-source remote sensing tree species identification method for unmanned aerial vehicle | |
| CN113592882A (en) | Crown extraction method based on multi-source remote sensing of unmanned aerial vehicle | |
| CN112381861A (en) | Forest land point cloud data registration and segmentation method based on foundation laser radar | |
| CN109815833A (en) | A kind of tea point recognition methods based on CCD Yu the feature extraction of three-dimensional laser sensor information fusion | |
| CN111340826A (en) | Single tree crown segmentation algorithm for aerial image based on superpixels and topological features | |
| Heenkenda et al. | Mangrove tree crown delineation from high-resolution imagery | |
| Hu et al. | A fast and simple method of building detection from LiDAR data based on scan line analysis | |
| CN111487643B (en) | Building detection method based on laser radar point cloud and near-infrared image | |
| Smits et al. | Individual tree identification using different LIDAR and optical imagery data processing methods | |
| Ferraz et al. | 3D segmentation of forest structure using a mean-shift based algorithm | |
| CN112669363A (en) | Urban green land three-dimensional green volume calculation method | |
| Hanapi et al. | A review on remote sensing-based method for tree detection and delineation | |
| Bulatov et al. | Automatic tree-crown detection in challenging scenarios | |
| CN105590313A (en) | Forest three- dimensional canopy morphological structure extraction method on the basis of active contour model | |
| Zheng et al. | Single shot multibox detector for urban plantation single tree detection and location with high-resolution remote sensing imagery | |
| CN105701856B (en) | A kind of vegetation extracting method and system | |
| Koc-San et al. | Plastic and glass greenhouses detection and delineation from worldview-2 satellite imagery | |
| CN112396615A (en) | Individual stumpage segmentation method and device based on adjustable hemispherical canopy camera | |
| CN117893924A (en) | Unmanned aerial vehicle laser radar point cloud single wood segmentation method based on tree crown shape | |
| Kim et al. | Generation of a DTM and building detection based on an MPF through integrating airborne lidar data and aerial images | |
| Hui et al. | Individual tree extraction from UAV lidar point clouds based on self-adaptive Mean shift segmentation | |
| CN115019190B (en) | Method for extracting complex ground type information of terrain breaking area based on aerial remote sensing platform |
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 |