CN106875407A - A kind of unmanned plane image crown canopy dividing method of combining form and marking of control - Google Patents
A kind of unmanned plane image crown canopy dividing method of combining form and marking of control Download PDFInfo
- Publication number
- CN106875407A CN106875407A CN201710070334.XA CN201710070334A CN106875407A CN 106875407 A CN106875407 A CN 106875407A CN 201710070334 A CN201710070334 A CN 201710070334A CN 106875407 A CN106875407 A CN 106875407A
- Authority
- CN
- China
- Prior art keywords
- image
- crown canopy
- crown
- canopy
- remote sensing
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 48
- 230000011218 segmentation Effects 0.000 claims abstract description 31
- 238000001914 filtration Methods 0.000 claims abstract description 20
- 238000006243 chemical reaction Methods 0.000 claims abstract description 8
- 239000002023 wood Substances 0.000 claims abstract description 8
- 230000006978 adaptation Effects 0.000 claims abstract description 7
- 230000008569 process Effects 0.000 claims description 9
- 230000000877 morphologic effect Effects 0.000 claims description 8
- 230000008859 change Effects 0.000 claims description 5
- 230000003628 erosive effect Effects 0.000 claims description 5
- 238000000926 separation method Methods 0.000 claims description 4
- 230000000007 visual effect Effects 0.000 claims description 4
- 230000005855 radiation Effects 0.000 claims description 3
- 230000003014 reinforcing effect Effects 0.000 claims description 3
- 238000005260 corrosion Methods 0.000 claims 1
- 230000007797 corrosion Effects 0.000 claims 1
- 238000007796 conventional method Methods 0.000 abstract description 3
- 238000001514 detection method Methods 0.000 description 6
- 238000011160 research Methods 0.000 description 3
- 238000012360 testing method Methods 0.000 description 3
- 230000002159 abnormal effect Effects 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 238000011835 investigation Methods 0.000 description 2
- 238000002372 labelling Methods 0.000 description 2
- 239000003550 marker Substances 0.000 description 2
- 238000011430 maximum method Methods 0.000 description 2
- 241000196324 Embryophyta Species 0.000 description 1
- 240000000731 Fagus sylvatica Species 0.000 description 1
- 235000010099 Fagus sylvatica Nutrition 0.000 description 1
- 240000007817 Olea europaea Species 0.000 description 1
- 230000032683 aging Effects 0.000 description 1
- 238000009395 breeding Methods 0.000 description 1
- 230000001488 breeding effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 230000010006 flight Effects 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000000638 solvent extraction Methods 0.000 description 1
- 230000003595 spectral effect Effects 0.000 description 1
- 238000001228 spectrum Methods 0.000 description 1
- 230000001502 supplementing effect Effects 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10032—Satellite or aerial image; Remote sensing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20036—Morphological image processing
-
- 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
- Image Processing (AREA)
- Image Analysis (AREA)
Abstract
The present invention relates to a kind of combining form and the unmanned plane image crown canopy dividing method of marking of control:The local remote sensing image in some width forest zones is obtained using unmanned plane, complete remote sensing image is obtained with ortho-rectification through inlaying;The disposal of gentle filter is carried out to green light band using gaussian filtering method;Crown canopy vertex position is detected from green light band using the local maximum searching method of self adaptation;Using morphology operations, changed by a force minimum and the crown canopy vertex position information of acquisition is imposed on image;For the true color remote sensing image of ortho-rectification, the only two-value image comprising crown canopy region and the non-class of crown canopy region two, external label of the non-crown canopy region that will be extracted as segmentation are obtained using ISODATA clustering algorithms;External label is imposed to by carrying out watershed transform segmentation on the image after force minimum conversion, accurate standing forest list wood crown canopy boundary information is obtained.The present invention efficiently solves the inaccurate problem of crown canopy boundary segmentation that conventional method is caused.
Description
Technical field
The present invention relates to a kind of combining form and the unmanned plane image crown canopy dividing method of marking of control.
Background technology
Tree crown obtains luminous energy and carries out the place of energy conversion as trees, and it is for research Forest Growth situation and dynamic
Change is significant.But due to the complexity and randomness of forest structure so that shape of tree-crown and boundary information are obtained
Take abnormal difficult.In recent years, gradually stepping up with the development of satellite remote sensing technology, especially Satellite Image Spatial Resolution,
The estimation precision of forest tree crown is improve, but is influenceed by factors such as weather, cycle, resolution ratio and costs so that acquisition
Remotely-sensed data far can not meet the demand of Forestry Investigation.Unmanned aerial vehicle remote sensing as emerging remote sensing technology, its distinctive motor-driven spirit
Active, ageing and low cost, data the advantage such as easily obtain and are increasingly becoming effective means of supplementing out economy of satellite remote sensing technology, and
Multiple fields are widely applied.Although the research for unmanned air vehicle technique is increasing, for visible ray unmanned plane shadow
As extracting the research of Forest Canopy structural information also in experimental stage, such as D í azvarela have evaluated common unmanned plane camera
The degree of reliability of image capturing crown parameters, and olive breeding field at the one of Spain Cordoba area is tried
Test, the RMSE of its hat width estimation has reached 0.28.The RGB shadow that Chianucci etc. is obtained using eBee unmanned flights system
Picture, and estimate the Forest Canopy coverage of beech woods with reference to LAB2 image classification methods, that its coefficient of determination R2 reaches 0.7 is left
It is right;Forest Canopy structure is entered using the cloud data of unmanned plane video generation in addition with Morgenroth, Mathews etc.
Row analysis, and achieve certain achievement.But the crown canopy dividing method of routine can cause the inaccurate problem of crown canopy boundary segmentation, this
Uncertainty is brought for the precision that unmanned aerial vehicle remote sensing obtains forest parameters.
The content of the invention
In view of this, it is an object of the invention to provide a kind of combining form and the unmanned plane image crown canopy of marking of control
Dividing method, efficiently solves the inaccurate problem of crown canopy boundary segmentation that conventional method is caused.
To achieve the above object, the present invention is adopted the following technical scheme that:A kind of combining form and marking of control nobody
Machine image crown canopy dividing method, it is characterised in that comprise the following steps:
Step S1:The local remote sensing image in some width forest zones is obtained using unmanned plane, to some width forest zone remote sensing figures
As being inlayed the complete remote sensing image that forest zone is obtained with ortho-rectification;
Step S2:The disposal of gentle filter is carried out to the green light band of complete remote sensing image using gaussian filtering method;
Step S3:Woods is detected from the green light band of complete remote sensing image using the local maximum searching method of self adaptation
Hat vertex position;
Step S4:Using morphology operations, the crown canopy vertex position information that will be obtained is changed by a force minimum
It is imposed on the green light band image after smothing filtering;
Step S5:For the complete remote sensing image that step S1 is obtained, obtained only comprising crown canopy using ISODATA clustering algorithms
Region and the two-value image of the non-class of crown canopy region two, external label of the non-crown canopy region that will be extracted as segmentation;
Step S6:Based on the result that step S4 and step S5 is obtained, external label is imposed to and is turned by force minimum
Watershed transform segmentation is carried out on image after changing, accurate standing forest list wood crown canopy boundary information is obtained.
Further, the local remote sensing image is true color image, and resolution ratio is between 0.05-0.20m.
Further, the specific method of the step S2 is as follows:Using a Gaussian distribution curve come to complete remote sensing shadow
The green light band of picture is processed, and reduces the noise level of image and the radiation intensity value on reinforcing crown canopy summit, and Filtering Formula is such as
Under:
In formula, G (i, j) is the i-th row, and image picture dot gaussian filtering result at j row, i, j are natural number, and σ is gaussian filtering
Mean square deviation, minimum crown canopy size carries out images filter treatment as window in σ values selection standing forest.
Further, the specific method of the step S3 is as follows:
Step S31:Potential crown canopy vertex position in sample ground is detected by a stationary window, potential crown canopy top is obtained
Point;
Step S32:Dynamic window using self adaptation is judged the potential crown canopy summit for obtaining, if current vertex
For the maximum of correspondence window area is then preserved, otherwise delete;Dynamic window size is by calculating eight, potential summit section side
It is determined to the change journey value of semivariance, the semivariance computing formula of its image picture element is:
In formula, γ (h) is experience en difference, xiBe the pixel position of image, h be two spaces of pixel separate away from
From Z (x) is correspondence image xiThe pixel value at place, N is the logarithm of the pixel pair under certain separation distance.
Further, the specific method of the step S4 is as follows:
Step S41:Inversion operation is carried out to the image f after filtering process, then the crown canopy summit for obtaining is carried out minimum
Value mark, obtains tag images, such as following formula:
In formula, fmIt is the tag images for obtaining, p is each pixel of image, tmaxIt is the maximum of image;
Step S42:By the calculating image f+1 and tag images f of pixelmBetween minimum, image is forced most
Small value conversion;
The step in, morphology calculate be expressed as:(f+1)∧fm, then using marking image fmTo (f+1) ∧ fmEnter
Row morphological erosion is rebuild, such as following formula:
In formula, fmpIt is the image after force minimum conversion,Description (f+1) ∧ fmIn tag images fmCover
Morphological erosion process of reconstruction under film.
Further, the specific method of the step S5 is as follows:Based on the complete remote sensing image for obtaining, using ISODATA
Non-supervised classification is classified, class categories number >=10 of setting, maximum iteration >=5;To the classification results for obtaining
Interpretation by visual observation is merged, and obtains the only two-value image comprising crown canopy region and non-this two class of crown canopy region, and will extract
The non-crown canopy region for going out is used as the external label split.
Further, watershed transform segmentation is as follows using formula in the step S6:
In formula, WaterShed () is watershed function;Mask is mask function;BOutMaskExternal label, i.e., non-crown canopy
Region, WcanopyIt is standing forest list wood crown canopy border.
The present invention has the advantages that compared with prior art:The present invention efficiently solves the woods that conventional method is caused
The hat inaccurate problem of boundary segmentation;Be conducive to quick effective extraction of forest tree crown information, be forest inventory investigation middle forest plant division
Number, the precise and high efficiency estimation offer of canopy density are provided powerful support for.
Brief description of the drawings
Fig. 1 is flow chart of the method for the present invention.
Fig. 2A is the unmanned plane image of the embodiment of the present invention one.
Fig. 2 B are the green light band filter results of the embodiment of the present invention one.
Fig. 2 C are the direct watershed segmentation results of the embodiment of the present invention one.
Fig. 2 D are the crown canopy summits that the embodiment of the present invention one is extracted using stationary window.
Fig. 2 E are that the embodiment of the present invention one removes exceptional value result using variable window.
Fig. 2 F are the non-crown canopy binary maps of crown canopy of the embodiment of the present invention one.
Fig. 2 G are the Morphological Reconstruction mark results of the embodiment of the present invention one.
Fig. 2 H are the inside and outside mark addition results of the embodiment of the present invention one.
Fig. 2 I are tag images segmentation results inside and outside the combination of the embodiment of the present invention one.
The unmanned plane image of Fig. 3 A embodiment of the present invention two.
The direct watershed segmentation result of Fig. 3 B embodiment of the present invention two.
The crown canopy fixed point that Fig. 3 C embodiment of the present invention two is extracted using stationary window.
The abnormal extreme point results of self-adapting window removal of Fig. 3 D embodiment of the present invention two.
Fig. 3 E embodiment of the present invention two forces crown canopy summit image.
The Morphological Reconstruction result of Fig. 3 F embodiment of the present invention two.
The non-crown canopy binary map of crown canopy of Fig. 3 G embodiment of the present invention two.
The direct watershed segmentation result of internal labeling image of Fig. 3 H embodiment of the present invention two.
Tag images segmentation result inside and outside the combination of Fig. 3 I embodiment of the present invention two.
Specific embodiment
Below in conjunction with the accompanying drawings and embodiment the present invention will be further described.
Fig. 1 is refer to, the present invention provides the unmanned plane image crown canopy dividing method of a kind of combining form and marking of control,
It is characterised in that it includes following steps:
Step S1:The local remote sensing shadow in forest zone of some width resolution ratio between 0.05-0.20m is obtained using unmanned plane
Some width forest zones remote sensing images are inlayed by picture and ortho-rectification obtains the complete remote sensing image in forest zone;The part
Remote sensing images at least should be the true color image comprising red, green, blue wave band, and image course and sidelapping rate >=80%,
Through inlaying the complete remote sensing image obtained with ortho-rectification without substantially splicing vestige.
Step S2:The disposal of gentle filter is carried out to the green light band of complete remote sensing image using gaussian filtering method;Specifically
Method is as follows:Processed come the green light band to complete remote sensing image using a Gaussian distribution curve (bell curve), subtracted
The noise level of few image and the radiation intensity value on reinforcing crown canopy summit, Filtering Formula are as follows:
In formula, G (i, j) is the i-th row, and image picture dot gaussian filtering result at j row, i, j are natural number, σ is gaussian filtering
Mean square deviation, minimum crown canopy size carries out images filter treatment as window in σ values selection standing forest.
Step S3:Woods is detected from the green light band of complete remote sensing image using the local maximum searching method of self adaptation
Hat vertex position;Specific method is as follows:
Step S31:First, potential crown canopy vertex position in sample ground is detected by a less stationary window, is obtained
Potential crown canopy summit;
Step S32:Dynamic window using self adaptation is judged the potential crown canopy summit for obtaining, if current vertex
For the maximum of correspondence window area is then preserved, otherwise delete;Dynamic window size is by calculating eight, potential summit section side
It is determined to the change journey value of semivariance, the semivariance computing formula of its image picture element is:
In formula, γ (h) is experience en difference, xiBe the pixel position of image, h be two spaces of pixel separate away from
From Z (x) is correspondence image xiThe pixel value at place, N is the logarithm of the pixel pair under certain separation distance.
Step S4:Using morphology operations, the crown canopy vertex position information that will be obtained is changed by a force minimum
It is imposed on the green light band image after smothing filtering;Specific method is as follows:
Step S41:First, inversion operation is carried out to the image f after the disposal of gentle filter, then the crown canopy summit to obtaining
Minimum mark is carried out, tag images, such as following formula are obtained:
In formula, fmIt is the tag images for obtaining, p is each pixel of image, tmaxIt is the maximum of image;
Step S42:Then, by the calculating image f+1 and tag images f of pixelmBetween minimum, image is carried out by force
Minimum value conversion processed;
The step in, morphology calculate be expressed as:(f+1)∧fm, then using marking image fmTo (f+1) ∧ fmEnter
Row morphological erosion is rebuild, such as following formula:
In formula, fmpIt is the image after force minimum conversion,Description (f+1) ∧ fmIn tag images fmCover
Morphological erosion process of reconstruction under film.
Step S5:For the complete remote sensing image that step S1 is obtained, obtained only comprising crown canopy using ISODATA clustering algorithms
Region and the two-value image of the non-class of crown canopy region two, external label of the non-crown canopy region that will be extracted as segmentation;Specific side
Method is as follows:Based on the complete remote sensing image for obtaining, classified using ISODATA non-supervised classifications, the classification class of setting
Not Shuo >=10, maximum iteration >=5;To obtain classification results interpretation is merged by visual observation, obtain only comprising crown canopy
Region and the two-value image of non-this two class of crown canopy region, and the non-crown canopy region that will be extracted as segmentation external label.
Step S6:Based on the result that step S4 and step S5 is obtained, external label is imposed to and is turned by force minimum
Watershed transform segmentation is carried out on image after changing, accurate standing forest list wood crown canopy boundary information is obtained.Watershed transform is split
It is as follows using formula:
In formula, WaterShed () is watershed function;Mask is mask function;BOutMaskExternal label, i.e., non-crown canopy
Region, WcanopyIt is standing forest list wood crown canopy border.
In order to allow those skilled in the art to be better understood from technical scheme, below in conjunction with two embodiments to this hair
It is bright to describe in detail.Wherein, the local remote sensing image that unmanned plane is obtained is RGB true color images, is carried out using PIX4D softwares
Pretreatment, through inlaying with after ortho-rectification, image resolution more 7cm.
Embodiment one:
Fig. 2A is 1 primary visible light image of sample ground, and sample ground 1 is coniferous forest sample ground, has isolated tree crown also to have overlapped
Tree crown.Fig. 2 B be green light band through maximum value filtering and the result of Gaussian smoothing filter, enhance the light of crown canopy and non-crown canopy
Spectral difference is different, and the spectrum reduced inside crown canopy is heterogeneous.Fig. 2 C are directly to carry out watershed to the green light band after filtering process
, there is the phenomenon of over-segmentation in segmentation.Because except crown canopy summit can also have partial noise value in image, and in image
The reason for there is road and vacant lot;
Fig. 2 D are the results that the detection of crown canopy summit is carried out using stationary window local maximum method, there is part crown canopy detection
To multiple summit problems;
Fig. 2 E are on the basis of stationary window testing result, woods to be carried out using variable window (self-adapting window) maximum value process
The result of crown point detection, it is found that eliminate the phenomenon that multiple summits occurs in part crown canopy;
Fig. 2 F are the crown canopy and non-crown canopy binary map obtained by unsupervised classification;
Fig. 2 G:It is the green light band for by Morphological Reconstruction force minimum transition treatment, now searches the tree for obtaining
Hat apex marker is imposed on image, that is, ensure that watershed segmentation can only be split according to these treetops mark;
Fig. 2 H are the results that non-crown canopy region exterior mark is increased in Fig. 2 G results;
Fig. 2 I are one tree crown of Polygons Representation of each closure with reference to external labeling watershed segmentation result in tree crown.It is logical
Cross and tree crown result of sketching is superimposed with raw video, it can be seen that the result that this algorithm is obtained is relatively preferable.
Embodiment two:
Fig. 3 A are 2 primary visible light images of sample ground, and sample ground 2 is broad-leaf forest sample ground.Fig. 2 B are the direct watershed of green light band
Segmentation result.There is the phenomenon of over-segmentation.Because except crown canopy summit can also have partial noise value, Yi Jiying in image
The reason for there is shrub ground as in;
Fig. 3 C are the results that the detection of crown canopy summit is carried out using stationary window local maximum method, there is part crown canopy detection
Summit problem is detected to multiple summit problems, and shrub on the ground;
Fig. 3 D are on the basis of stationary window testing result, woods to be carried out using variable window (self-adapting window) maximum value process
The result of crown point detection, it is found that eliminate part crown canopy and the phenomenon on multiple summits occur, but still push up with there is shrub
Point problem;
Fig. 3 E are the results that the tree crown apex marker for obtaining is imposed on green light band image;Fig. 3 F are by morphology
Reconstruct, as a result, the result ensures that watershed segmentation can only be split according to treetop mark.Additionally, in order to eliminate non-crown canopy area
Influence of the domain to crown canopy partitioning boundary, in addition it is also necessary to which mask process is carried out to segmentation result.Fig. 3 G are the woods that unsupervised classification is obtained
It is preced with non-crown canopy binary map.Fig. 3 H directly carry out watershed segmentation result for tree crown tag images, fail to influence with eliminating shrub;Figure
3I is the result after carrying out watershed segmentation on the basis of treetop mark and non-crown canopy mask mark, and the result for obtaining is relatively
It is good.
Two single wood crown canopy segmentation results in sample ground are carried out by the experiment segmentation result according to more than with visual interpretation result respectively
Contrast verification, its result is as shown in table 1.
The single ebon hat extraction accuracy analysis in the sample of table 1 ground
We can draw from table 1, and the crown canopy segmentation precision on two sample ground is all higher, the sample ground based on coniferous forest
Plot-01 reaches 94.54%, and the sample ground plot-02 based on broad-leaf forest then reaches 95.56%.
The foregoing is only presently preferred embodiments of the present invention, all impartial changes done according to scope of the present invention patent with
Modification, should all belong to covering scope of the invention.
Claims (7)
1. the unmanned plane image crown canopy dividing method of a kind of combining form and marking of control, it is characterised in that including following step
Suddenly:
Step S1:The local remote sensing image in some width forest zones is obtained using unmanned plane, some width forest zones remote sensing images are entered
Row inlays the complete remote sensing image that forest zone is obtained with ortho-rectification;
Step S2:The disposal of gentle filter is carried out to the green light band of complete remote sensing image using gaussian filtering method;
Step S3:Crown canopy top is detected from the green light band of complete remote sensing image using the local maximum searching method of self adaptation
Point position;
Step S4:Using morphology operations, changed by a force minimum and force the crown canopy vertex position information of acquisition
On green light band image after to smothing filtering;
Step S5:For the complete remote sensing image that step S1 is obtained, obtained only comprising crown canopy region using ISODATA clustering algorithms
With the two-value image of the non-class of crown canopy region two, external label of the non-crown canopy region that will be extracted as segmentation;
Step S6:Based on the result that step S4 and step S5 is obtained, external label is imposed to by after force minimum conversion
Image on carry out watershed transform segmentation, obtain accurate standing forest list wood crown canopy boundary information.
2. the unmanned plane image crown canopy dividing method of combining form according to claim 1 and marking of control, its feature
It is:The local remote sensing image is true color image, and resolution ratio is between 0.05-0.20m.
3. the unmanned plane image crown canopy dividing method of combining form according to claim 1 and marking of control, its feature
It is:The specific method of the step S2 is as follows:Using a Gaussian distribution curve come the green light band to complete remote sensing image
Processed, reduced the noise level of image and the radiation intensity value on reinforcing crown canopy summit, Filtering Formula is as follows:
In formula, G (i, j) is the i-th row, and image picture dot gaussian filtering result at j row, i, j are natural number, and σ is equal for gaussian filtering
Variance, minimum crown canopy size carries out images filter treatment as window in σ values selection standing forest.
4. the unmanned plane image crown canopy dividing method of combining form according to claim 1 and marking of control, its feature
It is:The specific method of the step S3 is as follows:
Step S31:Potential crown canopy vertex position in sample ground is detected by a stationary window, potential crown canopy summit is obtained;
Step S32:Dynamic window using self adaptation is judged the potential crown canopy summit for obtaining, if current vertex is right
Answer the maximum of window area then to preserve, otherwise delete;Dynamic window size is by calculating eight, potential summit profile direction half
The change journey value of variance is determined, and the semivariance computing formula of its image picture element is:
In formula, γ (h) is experience en difference, xiIt is the pixel position of image, h is two space separation distances of pixel, Z (x)
It is correspondence image xiThe pixel value at place, N is the logarithm of the pixel pair under certain separation distance.
5. the unmanned plane image crown canopy dividing method of combining form according to claim 1 and marking of control, its feature
It is:The specific method of the step S4 is as follows:
Step S41:Inversion operation is carried out to the image f after filtering process, minimum mark then is carried out to the crown canopy summit for obtaining
Note, obtains tag images, such as following formula:
In formula, fmIt is the tag images for obtaining, p is each pixel of image, tmaxIt is the maximum of image;
Step S42:By the calculating image f+1 and tag images f of pixelmBetween minimum, image is carried out force minimum turn
Change;
The step in, morphology calculate be expressed as:(f+1)∧fm, then using marking image fmTo (f+1) ∧ fmCarry out shape
State corrosion is rebuild, such as following formula:
In formula, fmpIt is the image after force minimum conversion,Description (f+1) ∧ fmIn tag images fmUnder mask
Morphological erosion process of reconstruction.
6. the unmanned plane image crown canopy dividing method of combining form according to claim 1 and marking of control, its feature
It is:The specific method of the step S5 is as follows:Based on the complete remote sensing image for obtaining, using ISODATA unsupervised classification sides
Method is classified, class categories number >=10 of setting, maximum iteration >=5;Classification results interpretation by visual observation to obtaining
Merge, obtain the only two-value image comprising crown canopy region and non-this two class of crown canopy region, and the non-crown canopy area that will be extracted
Domain is used as the external label split.
7. the unmanned plane image crown canopy dividing method of combining form according to claim 1 and marking of control, its feature
It is:Watershed transform segmentation is as follows using formula in the step S6:
In formula, WaterShed () is watershed function;Mask is mask function;BOutMaskIt is external label, Ji Fei crown canopies area
Domain, WcanopyIt is standing forest list wood crown canopy border.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710070334.XA CN106875407B (en) | 2017-02-09 | 2017-02-09 | Unmanned aerial vehicle image canopy segmentation method combining morphology and mark control |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710070334.XA CN106875407B (en) | 2017-02-09 | 2017-02-09 | Unmanned aerial vehicle image canopy segmentation method combining morphology and mark control |
Publications (2)
Publication Number | Publication Date |
---|---|
CN106875407A true CN106875407A (en) | 2017-06-20 |
CN106875407B CN106875407B (en) | 2020-06-12 |
Family
ID=59167303
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710070334.XA Expired - Fee Related CN106875407B (en) | 2017-02-09 | 2017-02-09 | Unmanned aerial vehicle image canopy segmentation method combining morphology and mark control |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106875407B (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107516068A (en) * | 2017-07-26 | 2017-12-26 | 福州大学 | A kind of method that single ebon hat is extracted in the high resolution image from unmanned plane |
CN109446983A (en) * | 2018-10-26 | 2019-03-08 | 福州大学 | A kind of coniferous forest felling accumulation evaluation method based on two phase unmanned plane images |
CN110189328A (en) * | 2019-06-11 | 2019-08-30 | 北华航天工业学院 | A kind of Remote sensing image processing system and its processing method |
CN111160236A (en) * | 2019-12-27 | 2020-05-15 | 北京林业大学 | Automatic dividing method for watershed canopy by combining forest region three-dimensional morphological filtering |
TWI749833B (en) * | 2020-10-29 | 2021-12-11 | 中華學校財團法人中華科技大學 | UAV sloping soil-rock watershed image identification method, system and application |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102096818A (en) * | 2011-01-31 | 2011-06-15 | 中国林业科学研究院资源信息研究所 | Object-oriented automatic extracting method and system for outline and parameter of remote sensing image crown |
CN103426200A (en) * | 2013-08-26 | 2013-12-04 | 天津大学 | Tree three-dimensional reconstruction method based on unmanned aerial vehicle aerial photo sequence image |
-
2017
- 2017-02-09 CN CN201710070334.XA patent/CN106875407B/en not_active Expired - Fee Related
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102096818A (en) * | 2011-01-31 | 2011-06-15 | 中国林业科学研究院资源信息研究所 | Object-oriented automatic extracting method and system for outline and parameter of remote sensing image crown |
CN103426200A (en) * | 2013-08-26 | 2013-12-04 | 天津大学 | Tree three-dimensional reconstruction method based on unmanned aerial vehicle aerial photo sequence image |
Non-Patent Citations (4)
Title |
---|
刘峰 等: "基于标记控制分水岭算法的树冠高程模型分割", 《中国农学通报》 * |
刘晓双 等: "高空间分辨率遥感的单木树冠自动提取方法与应用", 《浙江林学院学报》 * |
郭昱杉 等: "基于标记控制分水岭分割方法的高分辨率遥感影像单木树冠提取", 《地球信息科学学报》 * |
陈永富 等: "面向对象的退耕还林造林效果遥感特征提取技术研究", 《林业资源管理》 * |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107516068A (en) * | 2017-07-26 | 2017-12-26 | 福州大学 | A kind of method that single ebon hat is extracted in the high resolution image from unmanned plane |
CN109446983A (en) * | 2018-10-26 | 2019-03-08 | 福州大学 | A kind of coniferous forest felling accumulation evaluation method based on two phase unmanned plane images |
CN110189328A (en) * | 2019-06-11 | 2019-08-30 | 北华航天工业学院 | A kind of Remote sensing image processing system and its processing method |
CN111160236A (en) * | 2019-12-27 | 2020-05-15 | 北京林业大学 | Automatic dividing method for watershed canopy by combining forest region three-dimensional morphological filtering |
TWI749833B (en) * | 2020-10-29 | 2021-12-11 | 中華學校財團法人中華科技大學 | UAV sloping soil-rock watershed image identification method, system and application |
Also Published As
Publication number | Publication date |
---|---|
CN106875407B (en) | 2020-06-12 |
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 | |
Huang et al. | Building extraction from multi-source remote sensing images via deep deconvolution neural networks | |
Zhang et al. | Object-oriented shadow detection and removal from urban high-resolution remote sensing images | |
CN106875407A (en) | A kind of unmanned plane image crown canopy dividing method of combining form and marking of control | |
CN103049763B (en) | Context-constraint-based target identification method | |
Hormese et al. | Automated road extraction from high resolution satellite images | |
WO2018076138A1 (en) | Target detection method and apparatus based on large-scale high-resolution hyper-spectral image | |
Bunting et al. | Characterisation and mapping of forest communities by clustering individual tree crowns | |
CN101840581A (en) | Method for extracting profile of building from satellite remote sensing image | |
CN103646246A (en) | Decision tree model based multispectral remote sensing image river information extraction method | |
CN106294705A (en) | A kind of batch remote sensing image preprocess method | |
CN107992856B (en) | High-resolution remote sensing building shadow detection method under urban scene | |
CN103839267A (en) | Building extracting method based on morphological building indexes | |
Yang et al. | A multi-band watershed segmentation method for individual tree crown delineation from high resolution multispectral aerial image | |
CN104658003A (en) | Tongue image segmentation method and device | |
Qiao et al. | Urban shadow detection and classification using hyperspectral image | |
CN107516068A (en) | A kind of method that single ebon hat is extracted in the high resolution image from unmanned plane | |
CN105513060A (en) | Visual perception enlightening high-resolution remote-sensing image segmentation method | |
Huang et al. | Deep convolutional segmentation of remote sensing imagery: A simple and efficient alternative to stitching output labels | |
Huang et al. | Multi-feature combined for building shadow detection in GF-2 Images | |
Xu et al. | Instance segmentation of trees in urban areas from MLS point clouds using supervoxel contexts and graph-based optimization | |
CN102231190A (en) | Automatic extraction method for alluvial-proluvial fan information | |
CN107194405A (en) | A kind of method that interactive semi-automatic high-resolution remote sensing image building is extracted | |
CN114943869B (en) | Airport target detection method with enhanced style migration | |
CN104268836A (en) | Watershed segmentation mark point extraction method based on local area homogeneity indexes |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20200612 |