CN108020211A - A kind of method of unmanned plane aeroplane photography estimation instruction plant biomass - Google Patents

A kind of method of unmanned plane aeroplane photography estimation instruction plant biomass Download PDF

Info

Publication number
CN108020211A
CN108020211A CN201711248364.1A CN201711248364A CN108020211A CN 108020211 A CN108020211 A CN 108020211A CN 201711248364 A CN201711248364 A CN 201711248364A CN 108020211 A CN108020211 A CN 108020211A
Authority
CN
China
Prior art keywords
ground
instruction plant
model
biomass
vegetation index
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
Application number
CN201711248364.1A
Other languages
Chinese (zh)
Other versions
CN108020211B (en
Inventor
耿宇鹏
杨�琅
王彬
孙虎
***
张志明
姚扬
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Yunnan University YNU
Original Assignee
Yunnan University YNU
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Yunnan University YNU filed Critical Yunnan University YNU
Priority to CN201711248364.1A priority Critical patent/CN108020211B/en
Publication of CN108020211A publication Critical patent/CN108020211A/en
Application granted granted Critical
Publication of CN108020211B publication Critical patent/CN108020211B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
    • G01C11/00—Photogrammetry or videogrammetry, e.g. stereogrammetry; Photographic surveying
    • G01C11/04—Interpretation of pictures

Landscapes

  • Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Image Processing (AREA)

Abstract

The invention discloses a kind of method of unmanned plane aeroplane photography estimation instruction plant biomass, the image in region to be estimated is continuously shot by unmanned plane, while treats estimation region instruction plant biomass sample and is gathered on the spot and calculate biomass；Treat estimation area image to be handled, obtain high-resolution localized ground orthography to be estimated, and then establish region ground flora group height change phantom images and visible ray vegetation index phantom images to be estimated；Treat estimation localized ground orthography to classify, confirm the spatial distribution of instruction plant；Phytomass and phytobiocoenose height, the regression model of visible ray vegetation index are established, by regression model to having confirmed that the instruction plant biomass of spatial distribution is estimated；The method of the present invention can quickly estimate the biomass and spatial distribution of instruction plant, greatly improve work efficiency and quality, and more accurate, and cost is lower.

Description

A kind of method of unmanned plane aeroplane photography estimation instruction plant biomass
Technical field
The present invention relates to the instruction plant estimation of biomass technology in a kind of survey of natural resources, is specially a kind of unmanned plane The method that instruction plant biomass is estimated in aeroplane photography.
Background technology
Biotic intrusion is considered as to cause the second largest reason of global the loss of biodiversity, to local ecological environment, Economy, human health cause serious harm.How effective prevention and control invasive species, be the significant problem to receive much concern at present.Enter Exponential type expansion may be shown as by invading kind the population extension in new place, according to the principle of " early find, early eradicate ", invasive species Early warning is the committed step of prevention and control.Its particular content includes：For potential invasive species or invade but local The invasive species of distribution is quickly identified, with reference to its ecology and biological characteristics and environmental characteristic, assesses its occurrence injury Possibility, the scope and degree of harm, formulate feasible Prevention and control measures accordingly.Therefore, it is badly in need of developing a kind of skill at present Art quickly identifies the denizen population expanded, while rapid evaluation its distribution area, the extent of injury and generation Ratio etc..
The conventional method being monitored at present to the Distribution Pattern of instruction plant mainly includes two kinds：(1) artificial sample on the spot Square monitoring method；(2) satellite image monitoring method.Two methods respectively have advantage and disadvantage, and first method can clearly differentiate instruction plant With the difference of other plant, investigation result is more accurate, but time-consuming and laborious, high labor cost；Second method investigation is more simple Just, required time is short, and cost of labor is low, but the image resolution of the monitoring to instruction plant is poor, and investigation result error is larger.
UAV Aerial Surveying Technologies belong to Technology of low altitude remote sensing in the present invention, it is during image is obtained from atmospheric factor Interference, can obtain the data of taking photo by plane of Various Seasonal.With use cost is low, easy to operate, acquisition image speed is fast, ground distributor Resolution is high, can obtain the incomparable advantage of traditional remote sensing technology such as ground flora group height.It can be obtained with rich Rich textural characteristics and the aerial images of high spatial resolution.By using image processing platform, Remote Sensing Image Processing etc. Ground flora group height is obtained after software processing.And in visible channel, green vegetation is high in green channel reflectivity, It is low in feux rouges, blue channel reflectivity, vegetation and surrounding can be strengthened by the computing between green channel and feux rouges, blue channel The difference of atural object, therefore visible ray vegetation index can be established based on visible channel.This technology is to instruction plant biomass Estimation has the accuracy and feasibility of higher.
The content of the invention
In view of the problems of the existing technology, the present invention provides one kind to utilize unmanned plane aeroplane photography estimation instruction plant The method of biomass；This method, as Aerial Photography platform, region to be estimated is continuously shot by unmanned plane using unmanned plane Image, while treat estimation region instruction plant biomass sample and gathered on the spot and calculate sampled point instruction plant biology Amount, measures sampled point GPS location；
Estimation area image is treated using digital photogrammetry technology to be handled, and obtains high-resolution localized ground to be estimated Orthography, and then establish region ground flora group height change phantom images and visible ray vegetation index model shadow to be estimated Picture；Treat estimation localized ground orthography to classify, confirm the spatial distribution of instruction plant, obtain instruction plant classification shadow As model；
Using the image processing function of GIS software, make ground flora group height change phantom images and visible ray Vegetation index phantom images resolution ratio is identical, the sampling point position of instruction plant biomass sample with corresponding to region to be estimated In table phytobiocoenose height change phantom images and visible ray vegetation index phantom images, each sampled point is extracted in model shadow Ground flora group height value and visible ray vegetation index value as in；According to sampled point instruction plant biomass, ground flora Group's height value and visible ray vegetation index value, establish instruction plant biomass and phytobiocoenose height, visible ray vegetation index Regression model, by regression model to having confirmed that the instruction plant biomass of spatial distribution is estimated.
The time of selected instruction plant Image Acquisition is the florescence because the plant in the florescence on image classification feature compared with To be obvious, it is easy to distinguish with other ground classes on same image.
Wherein described high-resolution ground orthography acquisition methods are as follows：By carrying out matter to unmanned plane aerial photography image Amount optimization and GPS track information importing processing, using exercise recovery structure algorithm and various visual angles stereo reconstruction method, carry out image Splicing, generates sparse cloud data, inputs the latitude and longitude coordinates and elevation of ground control point, introduces control point, and determining will The coordinate system of orthography is generated, then sparse cloud is encrypted, forms dense point cloud, then based on dense point cloud data Triangulation Network Model is generated, cloud data all on image is in same plane with Triangulation Network Model as reference, is generated Ground orthography.
Since high-resolution ground orthography is ground flora group height change phantom images and visible ray vegetation The basic data of exponential model video generation, therefore, the generation of high-resolution ground orthography are to be modeled analysis extremely An important step is closed, aerial stereo images data is imported in image processing software Lightroom first the parameter of image is adjusted Section, improves image quality, then extracts the aerial survey GPS track information in the course line track recorder of UAV flight, and The image after quality treatment is imported, the corresponding GPS information for mutually matching image in the same time, corresponding GPS letters are assigned to all images Breath, the exercise recovery structure algorithm and various visual angles carried in software AgisoftPhotoscan is utilized by the image data after processing Stereo reconstruction algorithm, to image carry out splicing, generate sparse cloud data, input ground control point latitude and longitude coordinates and Elevation, introduces control point, determines that the coordinate system of orthography will be generated, allows the cloud data of all generations to be in same coordinate Under system, then sparse cloud is encrypted, forms dense point cloud, then Triangulation Network Model is generated based on dense point cloud data, with Triangulation Network Model as reference, makes cloud data all on image be in same plane, ultimately produces high-resolutionly Face orthography.
The area ground flora group height change phantom images acquisition methods to be estimated are as follows：The dense of acquisition will be encrypted Cloud data generates digital surface model DSM, and the ground variation model on basis is then sorted out using dense point cloud, then numeral Surface model DSM generates earth's surface variation model DEM with the ground variation model on basis using interpolation method, by digital surface model DSM and earth's surface variation model DEM superpositions subtract each other to obtain ground flora group height change MODEL C HM, that is, with obtaining area to be estimated Table phytobiocoenose height change phantom images.
The dense point cloud data obtained using software AgisoftPhotoscan encryptions are specially generated into digital surface mould Type（Digital Surface Model, DSM）, then using dense point cloud sort out basis ground variation model, then DSM generates earth's surface variation model DEM with the ground variation model on basis using interpolation method, utilizes the shadow of soft SA GA-GIS afterwards Subtract each other to obtain ground flora group height as digital surface model DSM and earth's surface variation model DEM is superimposed by data operation function Variation model (Canopy Height Model, CHM), that is, obtain area ground flora group height change model shadow to be estimated Picture.
Area's visible ray vegetation index phantom images acquisition methods to be estimated are as follows：Utilize Remote Sensing Image Processing （Software ENVI）Middle vegetation index model foundation function, estimation area ground orthography is treated using visible ray vegetation index formula Visible light wave range computing is carried out, generates visible ray vegetation index（VDVI）Phantom images.
The estimation localized ground orthography for the treatment of is classified, and confirms the method for spatial distribution of instruction plant such as Under：Localized ground orthography to be estimated is imported in software eCognition Developer, using multi-scale division algorithm, Multi-scale division, and the feux rouges of orthography visible ray, green light and blue light ripple carry out image based on the pixel layer of orthography Section both participates in segmentation, and segmentation scale parameter is arranged to 150-200；It is then based on the image split, instruction plant and non-is set Two kinds of ground class standards of instruction plant, and ground class description is write, different land types feature samples then are chosen on segmentation image, finally Classification image is generated with algorithm classification, has obtained distribution situation of the instruction plant in region to be estimated, most Export estimation region class Image model afterwards.
The instruction plant biomass sample to be estimated acquisition method on the spot：First according to instruction plant in selection area Growing way is high, in, short standard with determining several representative samples, sample area is 1 × 1 m2（The gross area is 1m2）, enter Invade phytomass and be collected as ground biomass, with then recording sample center GPS position information, ensure to choose sampling point and unmanned plane It is corresponding to splice image；The biomass of instruction plant is calculated using oven drying method, regression model is used it for and establishes analysis.
The instruction plant biomass and phytobiocoenose height, the regression model foundation of visible ray vegetation index and biomass Estimating and measuring method is as follows：It will be extracted from ground flora group height change phantom images and visible ray vegetation index phantom images The ground flora group height value and visible ray vegetation index value corresponding to sampling point position, established by software SPSS visible Light vegetation index and the function of a single variable relation of ground flora group height；SPSS is recycled to establish sampled point biomass and visible ray The binary linearity model of vegetation index, ground flora group height；Function of a single variable relation is substituted into binary linearity model and is drawn The generic function relational expression of biomass and visible ray vegetation index；The post processing of generic function relational expression is finally keyed in software ENVI can See that light vegetation index phantom images obtain biomass spatial model, realize instruction plant estimation of biomass.
Wherein function of a single variable relation y1=A1x1+C1(y1For ground flora group height, x1For VDVI values), biomass is with planting By index, the binary linearity model y of ground flora group height2=A2y1+B2x1+C2(wherein y2For biomass, y1Planted for earth's surface Thing group height, x1For VDVI values).By the ground flora group height y of binary linearity model1With the vegetation of function of a single variable relation Index y1It is replaced, replacement formula is：y2=A2（A1x1+C1）+B2x1+C2, the total of biomass and vegetation index is drawn after integration Functional relation；Generic function relational expression post processing vegetation index phantom images are keyed in software ENVI and obtain biomass spatial mode Type, is finally overlapped estimation region class Image model and biomass spatial model, realizes that instruction plant biomass distribution is estimated Survey.
The features of the present invention and advantage are：Biomass and phytobiocoenose height, the fitting of vegetation index bivariate, precision is more Height, applicability are wider；Unmanned plane can be perfectly suitable for the estimation of biomass of mesoscale regional extent, easy to spread；
The present invention confirms the spatial distribution of instruction plant by unmanned plane audio and video products, utilizes vegetation index VDVI and corresponding sample prescription The dependency relation of biomass and height average, estimates the biomass in instruction plant spatial distribution；Its estimation result obtained There is very strong correlation with actual biomass.
Brief description of the drawings
Fig. 1 is the high-resolution ground orthography model schematic of processing generation；
Fig. 2 is the visible ray vegetation index of processing generation（VDVI）Image model schematic diagram；
Fig. 3 is the broken line difference schematic diagram of estimation models check analysis；
Fig. 4 is estimation region class Image model schematic diagram；
Fig. 5 is estimation area's Biomass Models schematic diagram.
Embodiment
In order to which the technical features, objects and effects of the present invention are more clearly understood, below in conjunction with the accompanying drawings and implement The present invention will be further described for example, but the present invention is not limited to the content, and method is unless otherwise specified in embodiment It is conventional method.
Embodiment 1：
The unmanned plane aeroplane photography remote sensing platform and parameter used in the present embodiment is as follows：
The camera sensor maximum pixel of UAV flight is 7360 × 4912, and fixed aperture size is f/4.0, and ISO values are 100, shutter speed 1/1000S, and camera parameter compensation geometric distortion is called, lens parameters are fixed focal length 35mm；Nobody Machine uses big six rotor flying platforms of boundary M600 Pro, and joins software DJI Assistan2 calibrations using adjusting；Its dead end GPS makes With HOLUX M241-A track recorders.
For the present embodiment in order to obtain effective data, data acquisition carries out integrated planning design to aerial mission before, with Ensure the validity of data acquisition and the security of flight.
50% is should be higher than that in view of transverse overlap as defined in flight mapping, and sidelapping rate should be higher than that 20%, and gather Image data processing after, altitude information, the visible ray vegetation index of acquisition（VDVI）Data and the sampling point of actual measurement biomass Data correspond, and corresponding area is as close as possible, thus it is 4m/ to gather image data in the present embodiment to keep flying speed S, flight flying height are 100m, calculate ground resolution by ground resolution GSD and flying height relational expression, relational expression is as follows：
In formula, a is pixel dimension, and GSD is ground resolution, and f is lens focus, and h is flying height.
By calculating, Pixel size is 4.8256 μm, and flight flying height is 100m, and calculating ground resolution according to formula is 1.4cm。
The present embodiment region to be estimated is located at the Hua Nian towns in Yuxi E Shan counties of the Yunnan Province west and south, its position is in east longitude Between 102 ° of 02'-102 ° of 08', 24 ° of 02'-24 ° of 15' of north latitude, its climate type belongs to subtropical monsoon climate, the invasion of estimation area Plant is based on tithonia, and its distribution is wider, and biomass is larger, and surrounding physical features is open, and artificial shelter is few, fits Conjunction is monitored it；Image Acquisition choosing is carried out in the tithonia florescence.Obtain instruction plant biomass data, altitude information and can See light vegetation index（VDVI）Modeling data of the data as estimation models, concrete operations are as follows：
1st, instruction plant biomass obtains：The image in area to be estimated is continuously shot by unmanned plane, during unmanned plane during flying, Instruction plant biomass sampling operation is unfolded at the same time, measurement parameter mainly includes instruction plant biomass, sampled point GPS location； Instruction plant collection method be：It is high, medium and low definite representative according to instruction plant tithonia growing way in selection area 30 samples, sample area is 1 × 1m2；Instruction plant biomass is collected as ground biomass, with then recording sample center GPS position information, wetland ground flora group height to be estimated can be corresponded to by ensureing the sampled point of instruction plant biomass sample In variation model image and visible ray vegetation index phantom images；The biomass of instruction plant is finally calculated using oven drying method, will The aerial part of the instruction plant of collection is placed in baking oven, and each sample constant temperature drying at 85 DEG C reaches constant weight to it, weighs And record numerical value（As shown in table 1）, use it for modeling analysis.
1 biomass statistics table of table
Number sample Tithonia growing way Biomass（kg/m2） Number sample Tithonia growing way Biomass（kg/m2）
1 It is low 1.0910 16 In 2.2867
2 It is low 1.2164 17 In 2.2432
3 It is low 1.2480 18 In 2.4287
4 It is low 1.1594 19 In 2.7413
5 It is low 1.5007 20 In 2.4892
6 It is low 1.4295 21 It is high 2.5394
7 It is low 1.7135 22 It is high 2.7513
8 It is low 1.7012 23 It is high 2.9843
9 It is low 1.8580 24 It is high 3.3229
10 It is low 1.6893 25 It is high 3.2150
11 In 2.0148 26 It is high 3.4824
12 In 1.9342 27 It is high 3.5975
13 In 2.1081 28 It is high 3.2694
14 In 2.0490 29 It is high 4.0985
15 In 2.1924 30 It is high 4.3452
2nd, high-resolution ground orthography obtains
Ground flora group height model and visible ray vegetation index（VDVI）The generation of model will be based on high-resolutionly Face orthography, thus the altitude information and VDVI data of modeling could be obtained.Area's ortho-image processing is studied in the present embodiment Method is as follows：
The image data of unmanned plane collection can be effected by environmental factors, in gatherer process, it may occur that cumulus covers the sun The situation of light, causes partial image under-exposed, therefore carries out tune to image using image processing software Lightroom and consider and handle Reason, makes photo reach suitable exposure, and all photos are under same exposure, are ensured excellent in all qualities of image Change；
Then the HOLUX M241-A tracks of software EZtour extraction UAV flights are carried in the present embodiment using day dead end GPS Aerial survey GPS track information in recorder, and the image after quality treatment is imported, the GPS information that corresponding same time matching is shone, The photograph for assigning GPS information imports in software and will directly form take photo by plane track and seat in the plane position.
Then the image for matching GPS information is imported in software AgisoftPhotoscan, utilizes the fortune built in software Dynamic structure algorithm and the various visual angles stereo reconstruction algorithm method recovered carries out splicing to image, first, generates sparse cloud data, It is Highest to set accuracy parameter, and other specification selection is given tacit consent to；Then the latitude and longitude coordinates and elevation of ground control point are inputted （As shown in table 2）, control point is introduced, four angles in area are generally studied in the selection at control point, it is therefore an objective to determine to generate just The coordinate system of projection picture；Then sparse cloud is encrypted, forms dense point cloud, setting mass parameter is High, other ginsengs Number selection acquiescence；Again based on dense point cloud data generation Triangulation Network Model, it is Height field to set surface type parameter, number Dense point cloud is selected according to source, accuracy parameter is High；Triangulation Network Model with as reference, make point cloud all on image finally Data are in same plane, generate orthography（As shown in Figure 1）, above processing is in software AgisoftPhotoscan Realize.
The latitude and longitude coordinates and table of altitude at 2 control point of table
# Label X/East Y/North Z/Altitude
Point1 102.205638 24.073308 1109.951
Point2 102.206218 24.073640 1119.528
Point3 102.206078 24.073120 1118.545
Point4 102.205515 24.072895 1113.263
# Total error
3rd, treat estimation localized ground orthography to classify, confirm the spatial distribution of instruction plant
Orthography is imported in software eCognition Developer, using the image partition tools built in software, is based on The pixel layer of orthography carries out multi-scale division to image, and the feux rouges of orthography visible ray, green light and blue wave band are equal Segmentation is participated in, segmentation scale parameter is arranged to 150；Be then based on the image split, set tithonia, culture, water body, 5 kinds of ground class standards of exposed soil and other plant, and for all ground class standard ground class is write according to the classifying rules of average and standard deviation Description, then chooses different land types feature samples on segmentation image, is finally generated point with algorithm classification Class image, has thus obtained distribution situation of the instruction plant in area to be estimated, finally with Erdas Imagine Images lattice Formula export estimation region class Image model（As shown in Figure 4）.
4th, instruction plant group altitude information obtains：The dense point cloud obtained using software AgisoftPhotoscan encryptions Data generate digital surface model（Digital Surface Model, DSM）, then basis is sorted out using dense point cloud Ground variation model, then DSM is utilized interpolation method generation earth's surface variation model DEM, Zhi Houli with the ground variation model on basis DSM and DEM superpositions are subtracted each other to obtain ground flora group height change mould with the image data calculation function of soft SA GA-GIS Type (Canopy Height Model, CHM), that is, obtain research area's apparent height variation model image, afterwards in apparent height The altitude information in the sampled point collection sampling point of biomass is corresponded on variation model image（As shown in table 3）, use it for estimating The modeling analysis of model.
3 instruction plant group altitude information table of table
。
5th, visible ray vegetation index（VDVI）The acquisition of data：The high-resolution ground orthography of generation is imported distant Feel in image processing platform ENVI, Band Math are selected in Basic Tools menus, it is defeated in expression formula dialog box is inputted Enter the band math expression formula of corresponding vegetation index VDVI：(2*float(b2)-( float(b1)+ float(b3)))/ (2* Float (b2)+(float (b1)+float (b3))), to avoid data from overflowing, input the data type selection floating-point of wave band Type.The wave band of remote sensing image is selected afterwards, wherein specified red spectral band is b1, green light band b2, blue wave band b3, output Vegetation index model.Then the contrast of image picture elements is changed by the brightness value of Enhance change image picture elements, selected Linear 2%, linear stretch is done to image DN Distribution values between 2%~98%, and most exceptional values are given up in stretching Fall, so as to improve picture quality, generate final visible ray vegetation index（VDVI）Phantom images（As shown in Figure 2）, Zhi Hou Visible ray vegetation（As shown in table 4）, use it for the modeling analysis of estimation models；
4 visible ray vegetation index of table（VDVI）Tables of data
Numbering VDVI extraction of values Numbering VDVI extraction of values
1 0.0690 16 0.1538
2 0.0745 17 0.2385
3 0.0383 18 0.2463
4 0.1071 19 0.2359
5 0.0926 20 0.2186
6 0.1563 21 0.2264
7 0.1156 22 0.1927
8 0.1454 23 0.1865
9 0.1682 24 0.2746
10 0.1517 25 0.2155
11 0.1158 26 0.2359
12 0.1765 27 0.2438
13 0.1235 28 0.2864
14 0.1236 29 0.2986
15 0.1859 30 0.3129
The estimation models of the present embodiment are established（Verified containing model）, concrete operations are as follows：
（1）Phytomass and phytobiocoenose height, visible ray vegetation index（VDVI）Regression model foundation：It will obtain The instruction plant group altitude information and visible ray vegetation index arrived（VDVI）Data are imported in mathematical statistics software I BM SPSS, Using the curve estimation function of software I BM SPSS, VDVI and ground flora group height model y are obtained after processing1=2.2952+ 6.3621x1（R2=0.7503, P ＜ 0.05, wherein y1For ground flora group height, x1For VDVI）；Further through software I BM Instruction plant biomass data, instruction plant group altitude information and the visible ray vegetation index that processing same SPSS obtains （VDVI）Data, obtain on biomass and vegetation index, the binary linearity model y of ground flora group height2=0.6476y1+ 6.7000x1-1.0811(R2=0.8264, P ＜ 0.05, wherein y2For biomass, y1For ground flora group height, x1For VDVI).Variable replacement is highly carried out to VDVI and ground flora group, the ground flora group height y of binary linearity model1 With the vegetation index y of function of a single variable relation1After being replaced, replacement formula is：
y2=0.6476×（2.2952+6.3621x1）+6.7000x1- 1.0811, obtain biomass and vegetation index after integrating Generic function relational expression y2=10.8201x1+0.4053。
（2）Next by the visible ray vegetation index of generation（VDVI）Model is imported in Remote Sensing Image Processing ENVI, Band Math are selected in Basic Tools menus, are then obtained after input variable in inputting expression formula dialog box is replaced Vegetation index and biomass generic function relational expression y=10.8201x1+ 0.4053, analyzed, obtained intuitively by ArcGis Reflect the Biomass Models figure of the biomass in this region to be estimated（As shown in Figure 5）, step 3 is estimated region class image mould Type is overlapped with Biomass Models figure, finally realizes the instruction plant phytomass distribution in area to be estimated in the present embodiment Estimate purpose；
（3）Estimation of biomass model is verified：30 tithonia biomass samples are randomly selected from whole sample ground, are answered for model Inspection；The results are shown in Table 5 for it.Using the biomass of model estimation with surveying in biomass, maximum relative error 14.09% is minimum Relative error is 3.23%, and mean error 8.19%, the precision that as shown by data is entirely tested is higher, and methodological science is feasible；In order to We generate broken line disparity map for expression estimation of biomass model directly perceived and actual measurement difference, as shown in Figure 3.
5 model of table rechecks result
The foregoing is merely the schematical embodiment of the present invention, the scope of the present invention is not limited to.Any ability The technical staff in domain, equivalent variations, modification and the combination made on the premise of the design of the present invention and principle is not departed from, should all Belong to the scope of protection of the invention.

Claims (7)

  1. A kind of 1. method of unmanned plane aeroplane photography estimation instruction plant biomass, it is characterised in that：Using unmanned plane as distant Feel photographic platform, the image in region to be estimated is continuously shot by unmanned plane, while treat estimation region instruction plant biomass Sample is gathered and calculates sampled point instruction plant biomass on the spot, measures sampled point GPS location；
    Estimation area image is treated using digital photogrammetry technology to be handled, and obtains high-resolution localized ground to be estimated Orthography, and then establish region ground flora group height change phantom images and visible ray vegetation index model shadow to be estimated Picture；Treat estimation localized ground orthography to classify, confirm the spatial distribution of instruction plant, obtain instruction plant classification shadow As model；
    Using the image processing function of GIS software, make ground flora group height change phantom images and visible ray Vegetation index phantom images resolution ratio is identical, the sampling point position of instruction plant biomass sample with corresponding to region to be estimated In table phytobiocoenose height change phantom images and visible ray vegetation index phantom images, each sampled point is extracted in model shadow Ground flora group height value and visible ray vegetation index value as in；According to sampled point instruction plant biomass, ground flora Group's height value and visible ray vegetation index value, establish instruction plant biomass and phytobiocoenose height, visible ray vegetation index Regression model, by regression model to having confirmed that the instruction plant biomass of spatial distribution is estimated.
  2. 2. the method for unmanned plane aeroplane photography estimation instruction plant biomass according to claim 1, it is characterised in that high The localized ground orthography acquisition methods to be estimated of resolution ratio are as follows：By to unmanned plane aerial photography image carry out quality optimization and The importing of GPS track information is handled, and using exercise recovery structure algorithm and various visual angles stereo reconstruction method, stitching portion is carried out to image Reason, generates sparse cloud data, inputs the latitude and longitude coordinates and elevation of ground control point, introduces control point, determines to generate The coordinate system of orthography, is then encrypted sparse cloud, forms dense point cloud, then based on dense point cloud data generation three Angle pessimistic concurrency control, makes cloud data all on image be in same plane with Triangulation Network Model as reference, and generation ground is just Projection picture.
  3. 3. the method for unmanned plane aeroplane photography estimation instruction plant biomass according to claim 2, it is characterised in that treat It is as follows to estimate region ground flora group height change phantom images acquisition methods：The dense point cloud data generation obtained will be encrypted Digital surface model DSM, then sorts out the ground variation model on basis using dense point cloud, then digital surface model DSM Ground variation model with basis is changed digital surface model DSM and earth's surface using interpolation method generation earth's surface variation model DEM Model DEM superpositions subtract each other to obtain ground flora group height change MODEL C HM, that is, it is high to obtain region ground flora group to be estimated Spend variation model image.
  4. 4. the method for unmanned plane aeroplane photography estimation instruction plant biomass according to claim 1, it is characterised in that treat It is as follows to estimate region visible ray vegetation index phantom images acquisition methods：Utilize vegetation index model in Remote Sensing Image Processing Function is established, treating estimation wetland ground orthography using visible ray vegetation index formula carries out visible light wave range computing, raw Into visible ray vegetation index phantom images.
  5. 5. the method for unmanned plane aeroplane photography estimation instruction plant biomass according to claim 1, it is characterised in that right Localized ground orthography to be estimated is classified, and the method for confirming the spatial distribution of instruction plant is as follows：Will region be estimated Ground orthography is imported in software eCognition Developer, using multi-scale division algorithm, based on orthography Pixel layer carries out multi-scale division to image, and the feux rouges of the visible ray in orthography, green light and blue wave band both participate in point Cut, segmentation scale parameter is arranged to 150-200；
    The image split is then based on, instruction plant and the ground class standard of non-intrusive plant are set, and writes ground class description, is connect And choose different land types feature samples on segmentation image, finally generate classification image with algorithm classification, Distribution situation of the instruction plant in region to be estimated has been obtained, has finally exported the image classified.
  6. 6. the method for unmanned plane aeroplane photography estimation instruction plant biomass according to claim 1, it is characterised in that enter Invade phytomass and phytobiocoenose height, the regression model foundation of visible ray vegetation index and estimation of biomass method are as follows： Correspond to sampling by what is extracted from ground flora group height change phantom images and visible ray vegetation index phantom images Point position ground flora group height value and visible ray vegetation index value, by software SPSS establish visible ray vegetation index with The function of a single variable relation of ground flora group height；Recycle SPSS establish sampled point biomass and visible ray vegetation index, The binary linearity model of table phytobiocoenose height；Will function of a single variable relation substitute into binary linearity model in draw biomass with it is visible The generic function relational expression of light vegetation index；Generic function relational expression post processing visible ray vegetation index is finally keyed in software ENVI Phantom images obtain biomass spatial model, and biomass spatial model and instruction plant classification Image model are overlapped, real Now instruction plant estimation of biomass in region to be estimated.
  7. 7. the method for unmanned plane aeroplane photography estimation instruction plant biomass according to claim 1, it is characterised in that treat Estimating region instruction plant biomass sample, acquisition method is as follows on the spot：According to instruction plant growing way determine plant it is high, in, it is short Standard, in selected wetland Region according to instruction plant it is high, in, short standard with determining several samples, sample area is 1 × 1 m2；Instruction plant biomass is collected as ground biomass.
CN201711248364.1A 2017-12-01 2017-12-01 Method for estimating biomass of invasive plants through aerial photography by unmanned aerial vehicle Active CN108020211B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201711248364.1A CN108020211B (en) 2017-12-01 2017-12-01 Method for estimating biomass of invasive plants through aerial photography by unmanned aerial vehicle

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201711248364.1A CN108020211B (en) 2017-12-01 2017-12-01 Method for estimating biomass of invasive plants through aerial photography by unmanned aerial vehicle

Publications (2)

Publication Number Publication Date
CN108020211A true CN108020211A (en) 2018-05-11
CN108020211B CN108020211B (en) 2020-07-07

Family

ID=62078354

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201711248364.1A Active CN108020211B (en) 2017-12-01 2017-12-01 Method for estimating biomass of invasive plants through aerial photography by unmanned aerial vehicle

Country Status (1)

Country Link
CN (1) CN108020211B (en)

Cited By (22)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109459392A (en) * 2018-11-06 2019-03-12 南京农业大学 A kind of rice the upperground part biomass estimating and measuring method based on unmanned plane multispectral image
CN110148116A (en) * 2019-04-12 2019-08-20 深圳大学 A kind of forest biomass evaluation method and its system
CN110470275A (en) * 2019-09-02 2019-11-19 长沙理工大学 A method of withered riverbed bed ripples morphological parameters are measured based on UAV aerial survey terrain data
CN110598619A (en) * 2019-09-06 2019-12-20 中国农业科学院农业资源与农业区划研究所 Method and system for identifying and counting fruit trees by using unmanned aerial vehicle images
CN110717413A (en) * 2019-09-24 2020-01-21 云南师范大学 Unmanned aerial vehicle visible light image-oriented water body information extraction method and device
CN110779498A (en) * 2019-09-19 2020-02-11 中国科学院测量与地球物理研究所 Shallow river water depth mapping method and system based on unmanned aerial vehicle multi-viewpoint photography
CN110823813A (en) * 2019-11-19 2020-02-21 深圳市智绘科技有限公司 Forest land ground biomass estimation method
CN111047566A (en) * 2019-12-04 2020-04-21 昆明市滇池高原湖泊研究院 Method for carrying out aquatic vegetation annual change statistics by unmanned aerial vehicle and multispectral satellite image
CN111242224A (en) * 2020-01-16 2020-06-05 贵州省草业研究所 Multi-source remote sensing data classification method based on unmanned aerial vehicle extraction classification sample points
CN112016484A (en) * 2020-08-31 2020-12-01 深圳市赛为智能股份有限公司 Plant disturbance evaluation method and device, computer equipment and storage medium
CN112418075A (en) * 2020-11-20 2021-02-26 北京艾尔思时代科技有限公司 Corn lodging region detection method and system based on canopy height model
CN112767628A (en) * 2020-12-18 2021-05-07 交通运输部水运科学研究所 Port foreign organism intrusion disaster tracking and monitoring system
CN113034871A (en) * 2021-03-12 2021-06-25 湖南省林业科学院 Early warning analysis system for foreign plant distribution pattern and propagation path
CN113039982A (en) * 2021-03-12 2021-06-29 湖南省林业科学院 Invasive plant biological control system based on functional character screening model
CN113063742A (en) * 2021-03-24 2021-07-02 和数科技(浙江)有限公司 Method and system for measuring vegetation biomass, storage medium and terminal
CN113591946A (en) * 2021-07-15 2021-11-02 广西大学 Phoenix-eye blue biomass monitoring platform
CN113781431A (en) * 2021-09-10 2021-12-10 福州大学 Green vision rate calculation method based on urban point cloud data
CN113936214A (en) * 2021-10-19 2022-01-14 桂林理工大学 Karst wetland vegetation community classification method based on fusion of aerospace remote sensing images
CN114241331A (en) * 2021-12-16 2022-03-25 中国科学院南京地理与湖泊研究所 Wetland reed aboveground biomass remote sensing modeling method taking UAV as ground and Sentinel-2 intermediary
CN115760885A (en) * 2022-11-09 2023-03-07 南京林业大学 High-canopy-density wetland forest parameter extraction method based on consumption-level unmanned aerial vehicle image
CN117456364A (en) * 2023-11-09 2024-01-26 中国农业科学院农业信息研究所 Grassland biomass estimation method and system based on SfM and grassland height factors
CN118015473A (en) * 2024-04-08 2024-05-10 山东科技大学 Multi-spectrum unmanned aerial vehicle-based field peanut biomass inversion method

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2014006907A (en) * 2013-06-25 2014-01-16 Shimizu Corp Ecological network evaluation method and ecological network evaluation system using the same
CN104656098A (en) * 2015-02-03 2015-05-27 南京林业大学 Method for inverting remote sensing forest biomass

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2014006907A (en) * 2013-06-25 2014-01-16 Shimizu Corp Ecological network evaluation method and ecological network evaluation system using the same
CN104656098A (en) * 2015-02-03 2015-05-27 南京林业大学 Method for inverting remote sensing forest biomass

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
周在明等: "基于SPOT6遥感影像的滩涂湿地入侵种互花米草植株高度的反演研究", 《海洋学报》 *

Cited By (34)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US11029251B2 (en) 2018-11-06 2021-06-08 Nanjing Agricultural University Method for estimating aboveground biomass of rice based on multi-spectral images of unmanned aerial vehicle
CN109459392B (en) * 2018-11-06 2019-06-14 南京农业大学 A kind of rice the upperground part biomass estimating and measuring method based on unmanned plane multispectral image
CN109459392A (en) * 2018-11-06 2019-03-12 南京农业大学 A kind of rice the upperground part biomass estimating and measuring method based on unmanned plane multispectral image
CN110148116A (en) * 2019-04-12 2019-08-20 深圳大学 A kind of forest biomass evaluation method and its system
CN110470275A (en) * 2019-09-02 2019-11-19 长沙理工大学 A method of withered riverbed bed ripples morphological parameters are measured based on UAV aerial survey terrain data
CN110598619A (en) * 2019-09-06 2019-12-20 中国农业科学院农业资源与农业区划研究所 Method and system for identifying and counting fruit trees by using unmanned aerial vehicle images
CN110598619B (en) * 2019-09-06 2023-04-07 中国农业科学院农业资源与农业区划研究所 Method and system for identifying and counting fruit trees by using unmanned aerial vehicle images
CN110779498A (en) * 2019-09-19 2020-02-11 中国科学院测量与地球物理研究所 Shallow river water depth mapping method and system based on unmanned aerial vehicle multi-viewpoint photography
CN110717413A (en) * 2019-09-24 2020-01-21 云南师范大学 Unmanned aerial vehicle visible light image-oriented water body information extraction method and device
CN110717413B (en) * 2019-09-24 2022-11-25 云南师范大学 Unmanned aerial vehicle visible light image-oriented water body information extraction method and device
CN110823813A (en) * 2019-11-19 2020-02-21 深圳市智绘科技有限公司 Forest land ground biomass estimation method
CN111047566A (en) * 2019-12-04 2020-04-21 昆明市滇池高原湖泊研究院 Method for carrying out aquatic vegetation annual change statistics by unmanned aerial vehicle and multispectral satellite image
CN111242224A (en) * 2020-01-16 2020-06-05 贵州省草业研究所 Multi-source remote sensing data classification method based on unmanned aerial vehicle extraction classification sample points
CN112016484B (en) * 2020-08-31 2024-04-05 深圳市赛为智能股份有限公司 Plant invasion evaluation method, plant invasion evaluation device, computer equipment and storage medium
CN112016484A (en) * 2020-08-31 2020-12-01 深圳市赛为智能股份有限公司 Plant disturbance evaluation method and device, computer equipment and storage medium
CN112418075A (en) * 2020-11-20 2021-02-26 北京艾尔思时代科技有限公司 Corn lodging region detection method and system based on canopy height model
CN112418075B (en) * 2020-11-20 2024-01-26 北京艾尔思时代科技有限公司 Corn lodging area detection method and system based on canopy height model
CN112767628A (en) * 2020-12-18 2021-05-07 交通运输部水运科学研究所 Port foreign organism intrusion disaster tracking and monitoring system
CN113034871B (en) * 2021-03-12 2023-02-07 湖南省林业科学院 Early warning analysis system for foreign plant distribution pattern and propagation path
CN113034871A (en) * 2021-03-12 2021-06-25 湖南省林业科学院 Early warning analysis system for foreign plant distribution pattern and propagation path
CN113039982A (en) * 2021-03-12 2021-06-29 湖南省林业科学院 Invasive plant biological control system based on functional character screening model
CN113063742A (en) * 2021-03-24 2021-07-02 和数科技(浙江)有限公司 Method and system for measuring vegetation biomass, storage medium and terminal
CN113591946A (en) * 2021-07-15 2021-11-02 广西大学 Phoenix-eye blue biomass monitoring platform
CN113781431B (en) * 2021-09-10 2023-09-29 福州大学 Green view rate calculation method based on urban point cloud data
CN113781431A (en) * 2021-09-10 2021-12-10 福州大学 Green vision rate calculation method based on urban point cloud data
CN113936214B (en) * 2021-10-19 2023-06-02 桂林理工大学 Karst wetland vegetation community classification method based on fusion of aerospace remote sensing images
CN113936214A (en) * 2021-10-19 2022-01-14 桂林理工大学 Karst wetland vegetation community classification method based on fusion of aerospace remote sensing images
CN114241331A (en) * 2021-12-16 2022-03-25 中国科学院南京地理与湖泊研究所 Wetland reed aboveground biomass remote sensing modeling method taking UAV as ground and Sentinel-2 intermediary
CN114241331B (en) * 2021-12-16 2024-04-05 中国科学院南京地理与湖泊研究所 Remote sensing modeling method for ground biomass of reed in wetland by taking UAV as ground and Septinel-2 medium
CN115760885A (en) * 2022-11-09 2023-03-07 南京林业大学 High-canopy-density wetland forest parameter extraction method based on consumption-level unmanned aerial vehicle image
CN115760885B (en) * 2022-11-09 2023-09-19 南京林业大学 High-closure-degree wetland forest parameter extraction method based on consumer-level unmanned aerial vehicle image
CN117456364A (en) * 2023-11-09 2024-01-26 中国农业科学院农业信息研究所 Grassland biomass estimation method and system based on SfM and grassland height factors
CN117456364B (en) * 2023-11-09 2024-04-26 中国农业科学院农业信息研究所 Grassland biomass estimation method and system based on SfM and grassland height factors
CN118015473A (en) * 2024-04-08 2024-05-10 山东科技大学 Multi-spectrum unmanned aerial vehicle-based field peanut biomass inversion method

Also Published As

Publication number Publication date
CN108020211B (en) 2020-07-07

Similar Documents

Publication Publication Date Title
CN108020211A (en) A kind of method of unmanned plane aeroplane photography estimation instruction plant biomass
CN108007438A (en) The estimating and measuring method of unmanned plane aeroplane photography remote sensing wetland plant biomass
CN107909607B (en) A kind of year regional vegetation coverage computational methods
Capolupo et al. A novel approach for detecting agricultural terraced landscapes from historical and contemporaneous photogrammetric aerial photos
CN104899448B (en) A kind of self-adapting compensation method of the static localization scheme of Ensemble Kalman Filter
Wang et al. Landscape-level vegetation classification and fractional woody and herbaceous vegetation cover estimation over the dryland ecosystems by unmanned aerial vehicle platform
He et al. Using LiDAR data to measure the 3D green biomass of Beijing urban forest in China
Berveglieri et al. Identification of successional stages and cover changes of tropical forest based on digital surface model analysis
Herrero-Huerta et al. Dense Canopy Height Model from a low-cost photogrammetric platform and LiDAR data
Selim et al. Semi-automatic tree detection from images of unmanned aerial vehicle using object-based image analysis method
CN112800827A (en) Hyperspectral image classification experimental method
Hu et al. Integrating CART algorithm and multi-source remote sensing data to estimate sub-pixel impervious surface coverage: a case study from Beijing Municipality, China
Zhang et al. Understory biomass measurement in a dense plantation forest based on drone-SfM data by a manual low-flying drone under the canopy
Hu et al. Soil moisture retrieval using convolutional neural networks: Application to passive microwave remote sensing
CN110188419A (en) A kind of method of the analysis of landscape ecological tale quale and quantitative planning and designing
Umarhadi et al. The Comparison of Canopy Density Measurement Using UAV and Hemispherical Photography for Remote Sensing Based Mapping
Salimi Kouchi et al. Fractional vegetation cover estimation in urban environments
Sari et al. Quality analysis of single tree object with obia and vegetation index from lapan surveillance aircraft multispectral data in urban area
Rocchini Resolution problems in calculating landscape metrics
Liu et al. A Novel Impervious Surface Extraction Method Based on Automatically Generating Training Samples From Multisource Remote Sensing Products: A Case Study of Wuhan City, China
Afriansyah et al. Image Mapping Detection of Green Areas Using Speed Up Robust Features
Loghin et al. Supervised classification and its repeatability for point clouds from dense VHR tri-stereo satellite image matching using machine learning
Weinke et al. Strategies for semi-automated habitat delineation and spatial change assessment in an Alpine environment
Ozkan et al. Estimation of structural diversity in urban forests based on spectral and textural properties derived from digital aerial images
Bai et al. Estimating the Aboveground Biomass of Urban Trees by Combining Optical and Lidar Data: A Case Study of Hengqin, Zhuhai, China

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