CN108195767A - Estuarine wetland denizen monitoring method - Google Patents
Estuarine wetland denizen monitoring method Download PDFInfo
- Publication number
- CN108195767A CN108195767A CN201711416470.6A CN201711416470A CN108195767A CN 108195767 A CN108195767 A CN 108195767A CN 201711416470 A CN201711416470 A CN 201711416470A CN 108195767 A CN108195767 A CN 108195767A
- Authority
- CN
- China
- Prior art keywords
- satellite
- data
- information
- aerial vehicle
- unmanned aerial
- 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
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N2021/1793—Remote sensing
- G01N2021/1797—Remote sensing in landscape, e.g. crops
Abstract
The present invention provides a kind of estuarine wetland denizen monitoring method, S1：Unmanned aerial vehicle platform and satellite remote sensing film source are selected according to monitoring requirements；S2：In the course line of Target monitoring area setting unmanned plane and corresponding flight parameter is set to unmanned plane；S3：Unmanned Aerial Vehicle Data acquisition tasks are set；S4：Unmanned Aerial Vehicle Data acquisition tasks are performed, obtain Unmanned Aerial Vehicle Data；S5：The geography information of Unmanned Aerial Vehicle Data with satellite data is matched, obtains pretreatment satellite photo；S6：Patch properties information in extraction pretreatment satellite photo；S7 obtains optimal separability characteristic index according to Patch properties information sifting；S8：Classified according to optimal separability characteristic index to satellite data and precision evaluation is carried out to result.The present invention provides a kind of estuarine wetland denizen monitoring method, is classified using remote sensing software and GIS software to invasive species vegetation, has the advantages that high timeliness, high-resolution, accuracy is high, objectivity is strong and at low cost.
Description
Technical field
The present invention relates to Ecological Environmental Evaluation field more particularly to a kind of estuarine wetland denizen monitoring methods.
Background technology
The space characteristics and its variation of research and monitoring strand/estuarine wetland vegetation, are to grasp Wetland Environment feature
A main tool, can be the protection of wetlands ecosystems biology, the evaluating of benthic environment quality, impaired habitat it is ecological extensive
The scientific basis and managerial integration provided again.With deepening continuously for globalization process, it is wet that biotic intrusion causes strand/river mouth
The drastically forfeiture of bio-diversity in the ground ecosystem, monitoring and management non-native plant distributions situation are badly in need of paying attention to.However, only
Instruction plant spatial distribution far can not be studied in landscape scale by conventional survey means.Luckily the crowd of remote sensing technology
More advantages provide new method to solve this problem.
Remote sensing technology has broad covered area, room and time scale is various, spectral information enriches, observes flexible and data to obtain
It takes and the advantages such as facilitates, it has also become most cost-effectiveness and efficient data acquisition means in Wetland Environment monitoring, wet
Very important role is play in ground planning, management and protection.Due to intertidal zone by biotic intrusion, tidal action, put
The influence for waiting human activities is herded, easily generates the smaller plant of area or atural object patch, its spatial distribution is accurately monitored and move
State changes, and key needs the remote sensing image data source of high spatial resolution.High spatial resolution data (VHR) are used more and more
In the spatial distribution for studying and monitoring instruction plant species, they include Pleiades, QuickBird, IKONOS, IRS-P5 etc.
Space platform and nearly all aerial images, their spatial resolution some be even up to l meters it is following.On the other hand, by
In intertidal zone difference plant species spectral signature similarity degree height, it is necessary to appropriate divide in certain phenological periods and spatially and spectrally
Data are obtained under resolution, separability is increased by phenology information.It has been shown that the opportunity of data acquisition plays important work
With because vegetation is more different usually from around for certain period plants in vegetation season.Using variation monitoring method, with background
Vegetation is compared, and the life cycle difference of certain species can be identified.Since plant species may be by more in certain phenological stages
It identifies well, the acquisition opportunity of high spatial resolution remote sensing data is particularly significant.However, the procurement cost of VHR satellite datas is very
Height, under normal circumstances, the acquisition of VHR satellite datas are constrained by the normal trace of cloud and satellite.The spirit of satellite air movement
Activity is not high, and needs plan ahead, and expense is sufficiently expensive, and can not ensure not hidden by cloud layer by uncertain weather forecast
Gear.Therefore, the selection for carrying out VHR data phenological stages at this stage is limited.
China's unmanned air vehicle technique is ripe day by day in recent years, and unmanned plane has portability, low cost, low-loss, repeatable profit
With, risk is small, be easy to deployment etc. advantages, for atural object monitor provide high flexibility.In addition, unmanned function obtains on mesoscale
Atural object space characteristics in itself or textural characteristics or location information between neighbouring atural object, these information is taken to breach field
The space-time limitation of operation, the sophisticated category for atural object provide more comprehensively information, and then Optimum Classification algorithm.However, planting
In terms of object monitoring, the image only obtained by unmanned plane is classified, and it is relatively low to be faced with unmanned plane image spectrum resolution ratio,
The shortcomings that complex process is irregularly resulted in the need for due to geometric distortion and radiation.
Invention content
Deficiency in for the above-mentioned prior art, the present invention provide a kind of estuarine wetland denizen monitoring method, using now
There is ripe remote sensing software and GIS software to classify invasive species vegetation, have high timeliness, high-resolution,
High information quantity, on-site inspection is efficient, accuracy is high, objectivity is strong and it is at low cost the advantages of.
To achieve these goals, the present invention provides a kind of estuarine wetland denizen monitoring method, including step：
S1：A unmanned aerial vehicle platform and a satellite remote sensing film source are selected according to monitoring requirements；
S2：At least course line of a unmanned plane in a Target monitoring area is set and corresponding flight is set to the unmanned plane
Parameter；
S3：One Unmanned Aerial Vehicle Data acquisition tasks are set according to the climatic data of Target monitoring area；
S4：The Unmanned Aerial Vehicle Data acquisition tasks are performed, obtain Unmanned Aerial Vehicle Data；
S5：By the geography information of the Unmanned Aerial Vehicle Data and the satellite data progress from satellite remote sensing film source acquisition
Match, obtain pretreatment satellite photo；
S6：Extract the Patch properties information in the pretreatment satellite photo；
S7：Optimal separability characteristic index is obtained according to the Patch properties information sifting；
S8：Classified according to the optimal separability characteristic index to the satellite data and precision is carried out to result and commented
Valency.
Preferably, in the S1 steps, the satellite remote sensing film source include Pleiades remote sensing high resolution images source,
QuickBird remote sensing high resolution image sources and IKONOS remote sensing high resolution images source.
Preferably, in the S2 steps, the flight parameter include course line, shooting height, shooting angle, flying speed and
Shooting interval；The course line, shooting height and shooting angle are set according to the satellite data of the satellite remote sensing film source；Root
The flying speed is set according to the course line；The shooting interval is set according to the flying speed, the unmanned aerial vehicle according to
The adjacent photo of the shooting interval shooting has image overlap area；Obtain the intertidal zone cross section of the Target monitoring area
Distance d.
Preferably, the S3 steps further comprise step：
S31：Obtain the tide table of the Target monitoring area；
S32：Elevation of high water, elevation of low water and the tide of the Target monitoring area spring season is obtained according to the tide table
When；
S33：According to the farthest flying distance D of the unmanned plane and intertidal zone cross-sectional distance d, determine it is described nobody
The shooting of machine performs the period, forms the Unmanned Aerial Vehicle Data acquisition tasks；
As D ＞ d, the shooting execution period of the unmanned plane was located at before the time of tide of the elevation of high water, institute
The shooting for stating unmanned plane performs the duration of period more than or equal to the two-way time that the unmanned plane is shot along the course line；
As D≤d, staff need to carry unmanned plane and be moved to farthest to fly from water distance apart from waterside from d ' less than described
Entered back into during row distance D the unmanned plane shooting perform the period, the unmanned plane shooting perform the period when grow up
In the sum of the two-way time shot equal to the unmanned plane along the course line and the two-way time of the staff.
Preferably, the S5 steps further comprise step：
S51：The satellite data is carried out to the pretreatment of Band fusion and geometric correction, obtains pretreatment remotely-sensed data；
S52：The Unmanned Aerial Vehicle Data is imported by GIS software；
S53：By the Unmanned Aerial Vehicle Data vector quantization, the Unmanned Aerial Vehicle Data after vector quantization includes picture data, first
GPS information and floristics information, each first vector point and first GPS information and the plant species of the picture data
Category information is associated with；
S54：The pretreatment remotely-sensed data includes satellite image data, the second GPS information, unknown floristics spectrum letter
Breath and spatial information；Using first GPS information and the consistency of second GPS information, by the satellite image data
Each second vector point and the picture data the first vector point described in associated each information match so that it is described not
Know floristics spectral information and the floristics information match.
Preferably, in the S6 steps, same plantation is obtained using the dividing function of eCognitionDeveloper softwares
Object range shared in the pretreatment satellite photo is formed according to floristics in the pretreatment satellite photo multiple
Patch；Using the classification feature of eCognitionDeveloper softwares the corresponding plant species is associated with for the patch assignment
Category information；Patch properties information is formed, the Patch properties information includes：Patch spectral information, patch shape information and patch
Location information.
Preferably, the S7 steps further comprise step：
Using the output function in the eCognitionDeveloper softwares, the system of the Patch properties information is exported
It counts；
By statistical data described in multivariate analysis of variance, filter out the optimal separability characteristic index and determine per a kind of
The threshold range of vegetation.
Preferably, the S8 steps further comprise step：
Sorting algorithm is established by the eCognitionDeveloper softwares based on the optimal separability characteristic index
Rule set；
Using the set of algorithms in the eCognitionDeveloper softwares supervised classification to the satellite data into
Row classification obtains a classification pixel set；
By the classification pixel set compared with a reference image metaset cooperation, the precision of the satellite data is evaluated.
The present invention makes it have following advantageous effect as a result of above technical scheme：
First, the present invention utilizes the space-time convenience of Unmanned Aerial Vehicle Data and the function of vegetation and plant species information can be accurately positioned,
With reference to the abundant spectrum of high spatial resolution remote sensing data and spatial texture information and the disposable of ripe classification software can be used
Property, a kind of accurate monitoring method for being appropriate for vegetation and plant species is provided, there is advantage in the selected invasive species of monitoring.This
Method is overcome in wetland Region only by high-definition remote sensing data monitoring, is limited by weather, tide, cost, can
It is few with data, it is difficult to the difficulties of invasion vegetation are accurately monitored by enough phenology or spectral information.Realize space/
Optimum organization and balance between spectrum/temporal resolution, cost, precision.
Secondly, the implementation of unmanned plane sampling also overcomes the difficulty in wetland sampling point sampling on the spot, and unmanned plane is easily operated, covers
Lid range is big, reduces time and the workload of artificial field work, improves the efficiency of on-site inspection, can not only obtain artificial sampling
The inaccessiable a large amount of number of samples amounts of institute, and the Patches information in its not obtainable mesoscale can also be obtained, it improves and divides
Class precision.
Description of the drawings
Fig. 1 is the flow chart of the estuarine wetland denizen monitoring method of the embodiment of the present invention.
Specific embodiment
Below according to attached drawing 1, presently preferred embodiments of the present invention is provided, and be described in detail, make to be better understood when this
Function, the feature of invention.
Referring to Fig. 1, a kind of estuarine wetland denizen monitoring method of the embodiment of the present invention, including step：
S1：A unmanned aerial vehicle platform and a satellite remote sensing film source are selected according to monitoring requirements.
In the present embodiment, 3 unmanned planes of Phantom of Liao great boundarys company (DJI Corporation) are selected, this is a kind of
It is small-sized, battery powered, carry high-definition camera, relatively cheap unmanned plane, it can set flight course planning, shooting angle automatically
Degree, height, shooting interval and flying speed；The photo of acquisition is integrated with built-in GPS geography information, and nothing can be carried out with satellite data
Difference matching.
Satellite remote sensing film source includes Pleiades remote sensing high resolution images source, QuickBird remote sensing high resolution images
Source and IKONOS remote sensing high resolution images source etc.；Other existing remote sensing high resolution image sources can also be used to its specific kind
Class is not limited.
In addition when selecting film source, appropriate phenological period film source is selected under optional film source as possible, selection monitoring as possible is planted
By species and surrounding vegetation otherness larger time in phenological period, such as budding time, florescence or withering period.
S2：The setting at least course line of a unmanned plane and the corresponding flight ginseng to unmanned plane setting in a Target monitoring area
Number.
Flight parameter includes course line, shooting height, shooting angle, flying speed and shooting interval；According to satellite remote sensing piece
Satellite data setting course line, shooting height and the shooting angle in source；Flying speed is set according to course line；It is set according to flying speed
Shooting interval, unmanned aerial vehicle have image overlap area according to the adjacent photo that shooting interval is shot.It can be obtained when setting course line
Obtain the intertidal zone cross-sectional distance d of Target monitoring area.
In the present embodiment, it can be set according to the heterogeneous feature of the remote sensing images landscape in the satellite data of satellite remote sensing film source
Put course line；According to the recognizable degree of species, Patch size sets shooting height in picture；It is clapped according to Vegetation canopy feature-set
Take the photograph angle；Flying speed is set according to the distance of each airline operation；Shooting interval is set according to flying speed, makes photo-overlap
And diagonal is respectively between picture altitude and 80% to the 85% of width, ensure to have between photo and photo it is enough overlapping,
Realize the uniform fold of sampling area.
S3：One Unmanned Aerial Vehicle Data acquisition tasks are set according to the climatic data of Target monitoring area, further comprise walking
Suddenly：
S31：Obtain the tide table of Target monitoring area；
S32：According to elevation of high water, elevation of low water and the time of tide of tide table acquisition Target monitoring area spring season, and according to
Elevation of high water, elevation of low water and the time of tide obtain the intertidal zone cross-sectional distance d of Target monitoring area；
S33：According to the farthest flying distance D of unmanned plane and intertidal zone cross-sectional distance d, determine that the shooting of unmanned plane performs
Period forms Unmanned Aerial Vehicle Data acquisition tasks；
As D ＞ d, the shooting execution period of unmanned plane was located at before the time of tide of elevation of high water, and the shooting of unmanned plane is held
The duration of row period is more than or equal to the two-way time that unmanned plane is shot along course line；
As D≤d, staff need to carry unmanned plane be moved to apart from waterside from water distance from d ' less than farthest flight away from
The shooting that unmanned plane is entered back into during from D performs the period, and the shooting of unmanned plane performs the duration of period and is more than or equal to unmanned plane
The sum of the two-way time shot along course line and the two-way time of staff.
S4：Unmanned Aerial Vehicle Data acquisition tasks are performed, obtain Unmanned Aerial Vehicle Data.
S5：The geography information of Unmanned Aerial Vehicle Data with the satellite data obtained from satellite remote sensing film source is matched, is obtained
Pre-process satellite photo；It further comprises step：
S51：Satellite data is carried out to the pretreatment of Band fusion and geometric correction, obtains pretreatment remotely-sensed data；
S52：Unmanned Aerial Vehicle Data is imported by GIS software；In the present embodiment, GIS software is adopted
With ArcGIS softwares；
S53：Pass through the Geotagged photos to point functions of the Data Management in ArcGIS softwares
For module by Unmanned Aerial Vehicle Data vector quantization, the Unmanned Aerial Vehicle Data after vector quantization includes picture data, the first GPS information and floristics
Information, each first vector point of picture data and the first GPS information and floristics information association.
S54：Pre-process remotely-sensed data include satellite image data, the second GPS information, unknown floristics spectral information and
Spatial information；Using the first GPS information and the consistency of the second GPS information, by each second vector point of satellite image data with
The first associated each information match of vector point of picture data so that unknown floristics spectral information and floristics information
Match.
By the step for can obtain in satellite image data certain vegetation title, spectral information and spatial information for putting.
Again by step S7, using Classification in Remote Sensing Image software, by patch where vegetation and plant species sampling point known in satellite image data, sampling point
Spectral information and spatial information feature extraction come out, and are analyzed.
Comprising pretreatment to high spatial resolution remote sensing data, i.e. satellite data in this step, initial data need into
Row Band fusion and geometric correction can just become the accurate data available of high-resolution and position.These steps can be by existing
There are the Band fusion of remote sensing software and geometric correction Implement of Function Module.
Since high spatial resolution remote sensing data includes multi light spectrum hands (such as P stars are 2m resolution ratio) and panchromatic wave-band (such as P
Star is 0.5m resolution ratio).Multi light spectrum hands product provides the object spectrum information than panchromatic wave-band product more horn of plenty, can be formed
Color composite photograph, the advantageous identification for using atural object, but spatial resolution is low compared with panchromatic wave-band.Panchromatic wave-band single band, on the diagram
Display is gray scale picture, General Spatial high resolution, but can not explicitly look for coloured silk.By panchromatic wave-band and multi light spectrum hands image
Fusion treatment obtains the high-resolution of existing panchromatic image, and has the image of the colour information of multiband image.
When geometric correction generally refers to be imaged to correct and eliminate remote sensing image by a series of mathematical model, due to flying
The influence of the factors such as posture, height, speed and the earth rotation of row device causes image abnormal relative to ground target generation geometry
Change, this distortion show as pixel and extruding, distortion, stretching and offset etc. occur relative to the physical location of ground target, for
Geometric correction is just named in the error correction that geometric distortion carries out.Geometric accurate correction：The geometric correction carried out using control point, it is to use
A kind of mathematical model carrys out the geometric distortion process of approximate description remote sensing images, and using distortion remote sensing images and standard map it
Between some corresponding points (control point data to) acquire this geometric distortion model, then this model is utilized to carry out geometric distortion
Correction.
The built-in GPS module of unmanned plane, if after 3 or more satellite-signals can be received simultaneously, just can more it is fast more accurately
GPS positioning data are obtained, and are recorded in the EXIF information of photo.
S6：Multiple patches are formed in satellite photo is pre-processed according to floristics, and form the Patch properties letter of patch
Breath.
Same plant is obtained in satellite photo is pre-processed using the dividing function of eCognitionDeveloper softwares
Shared range forms multiple patches according to floristics in satellite photo is pre-processed；Utilize eCognitionDeveloper
The classification feature of software is associated with corresponding floristics information for patch assignment；Form Patch properties information, Patch properties information
Including：Patch spectral information, patch shape information and plaque location information.
S7：Optimal separability characteristic index is obtained according to Patch properties information sifting.
Using output (Export) function in eCognitionDeveloper softwares, the system of Patch properties information is exported
It counts；By multivariate analysis of variance statistical data, filter out optimal separability characteristic index and determine the threshold of every a kind of vegetation
It is worth range.
In the present embodiment, in order to differentiate the whether significant difference of the different characteristic information of different vegetation classification patches, lead to
Multivariate analysis of variance statistical data is crossed, in the present embodiment, can be by the Implement of Function Module operation of various statistical softwares
Multivariate analysis of variance (analysis is carried out by SPSS softwares>General linear model>Multivariable) it realizes, to different type plant spot
The sample average otherness of block's attribute carries out significance test, filters out optimal separability characteristic index.
In statistical analysis, can obtain mean value under the different characteristic parameter of different vegetation types, standard error and
95% fiducial interval range, these ranges constitute the threshold value model per a kind of vegetation under the characteristic parameter with significant difference
It encloses.
S8：Classified according to optimal separability characteristic index to satellite data and precision evaluation is carried out to result.
First, sorting algorithm is established by eCognitionDeveloper softwares based on optimal separability characteristic index to advise
Then collect；In the present embodiment, sorting algorithm rule set is established using eCognition Developer softwares, it is only necessary to first establish institute
Some atural object item names, the supervised classification algorithm then used under every a kind of atural object；Such as minimum neighbouring sorting algorithm；Again
Determining optimization characteristic parameter is selected into addition per under a kind of atural object classification algorithm.
Secondly, satellite data is carried out using the supervised classification of the set of algorithms in eCognitionDeveloper softwares
Classification obtains a classification pixel set；Supervised classification is to be instructed using establishing statistics recognition function as theoretical foundation, according to typical sample
Practice the technology that method is classified, i.e., by selecting characteristic parameter, characteristic parameter is obtained in the sample provided according to known training center
As decision rule, discriminant function is established with the image classification to each image progress to be sorted, is a kind of method of pattern-recognition.
In supervised classification, the selection of characteristic parameter is extremely important, and therefore, the purpose of this step is exactly that analysis determines optimal distinguish
The characteristic parameter of different vegetation types；
Finally, classification pixel set is evaluated into the precision of satellite data compared with a reference image metaset cooperation.
Now by the method applied in the present invention for estuary area Wetlands Monitoring instruction plant Spartina alterniflora applies：
Research ground is located at the Dong Tan of island-Chongming Island most the east of entrance of Changjiang River maximum, which is located in Chongming Dongtan country
Grade nature reserve area, is a typical strand/estuarine wetland.For the needs in soil, Spartina alterniflora starts to introduce for 1979
China, be a kind of salt tolerant for adapting to the growth of seabeach intertidal zone, it is resistance to flood plant, be mainly used for protecting beach bank protection, created land with silt and improvement
Soil etc..Spartina alterniflora is subordinate to grass family spartina category, is perennial tall and big herbaceous plant, robust plant and tall and straight, average plant height
About 1.5m.Since Spartina alterniflora has a series of mechanism for being conducive to the existence of its population and diffusion, inhibit other plant growths,
Make shellfish activity difficulty or even meeting death by suffocation in intensive spartina grass beach, threaten the food source of fish, birds, drop
The bio-diversity of low beach, the serious natural equilibrium for destroying Chongming Dongtan ecological sensitive areas, produces locality serious life
State changes the food web and trophic structure of the entire ecosystem with evolutionary consequences, and original many fish and migratory bird is caused to lose
Nutrition supports, and faces a danger.
Chongming Dongtan phytobiocoenose has the features such as dominant plant type is few, and coenotype is simple, is conducive to remote sensing monitoring
Research.Although Chongming Dongtan shares about 96 kinds of higher plant, sociales be mainly reed (Phragmitesaustralis),
Spartina alterniflora (Spartinaalterniflora), rough leaf sedge (Carexscabrifolia), extra large scirpus scirpus
(Scirpusmariqueter).In geographical distribution, vegetation divides zoning with certain, i.e., first by the light beach of low tide
Evolve as using extra large scirpus scirpus as sociales, have scirpus triqueter, rough leaf sedge group, secondly drilled in middle tidal level and high tide level
For for the group based on reed or Spartina alterniflora, and form the situation vied each other.Reed and Spartina alterniflora can usually give birth to
Grow to 1.5~2.0m, and scirpus triqueter, rough leaf sedge, extra large scirpus scirpus growing height be about 0.3~0.7m.
The first step：Select unmanned aerial vehicle platform and satellite remote sensing film source.
In the present embodiment, selection be great Jiang companies (DJI Corporation) 3 unmanned planes of Phantom, for defending
Star remote sensing film source has selected the Pleiades star remote sensing high resolution images source of mainstream, spatial resolution 0.5m, according on the spot
Observation, it is poor that the dominant species such as reed, Spartina alterniflora, extra large scirpus scirpus, the rough leaf sedge of Chongming Dongtan show certain phenology
Different, there is tandem in the especially seeding stage in spring and autumn withering period, according to the quality (angle, cloud amount) of optional image, finally
The remote sensing image in 10 months 2015 has been selected as target decomposition image.The period, reed mid-October are its full-bloom stage；
And 7 the end of month of Spartina alterniflora is full-bloom stage, mid-October is bears seeds the phase；Rough leaf sedge, extra large scirpus scirpus mid-October
Start to wither.
Second step：In the course line of Target monitoring area setting unmanned plane and corresponding flight parameter is set to unmanned plane.
During unmanned plane during flying task, course line is set according to the heterogeneous feature of remote sensing images landscape, is set from south to north
Three flight erect-positions are put, each erect-position sets three course lines to three directions, and 9 course lines, make all course lines more uniform altogether
The entire research area's range of ground distribution；Shooting height is set according to the recognizable degree (Patch size) of species, according to field on the spot
Investigation experience, most of medium and small Plant patch diameter range is about 10~30m or so, is tested according to practical flight, flying height
It is suitable with patch diameter range, shoot the higher 15~20m of recognizable degree of photo；According to Vegetation canopy height and vegetation type
Shooting angle is set, Dongtan wetland dominant plant is herbaceous plant, and blade canopy is narrower, and the big multiregion of plant height is on 0.5~2m left sides
The right side, if just penetrating vertical view shooting is difficult to differentiate between vegetation type, shooting angle is set as 30 °~45 ° of the elevation angle, can preferably differentiate in photo
Floristics；Flying speed is set according to the distance of each airline operation, since every airline distance is in 1km~2km or so, root
According to test, flying speed is set as 10m/s, and shooting interval 2s/ can make photo-overlap and diagonal respectively in picture altitude and width
Between 80% to the 85% of degree, ensure to have between photo and photo enough overlapping, realize the uniform fold of sampling area.
Third walks：One Unmanned Aerial Vehicle Data acquisition tasks are set according to the climatic data of Target monitoring area.
The tide table of the investigated wetland Region of inquiry, learn the elevation of high water of the spring season in the region, elevation of low water and
The time of tide；According to the elevation of high water of spring season, elevation of low water and the time of tide and the elevation in region is logged in and withdrawn, determines operating personnel
The executable period that the time and unmanned plane low altitude aircraft device for logging in and withdrawing take photo by plane.
4th step：Unmanned Aerial Vehicle Data acquisition tasks are performed, obtain Unmanned Aerial Vehicle Data.
5th step：By the geography information of Unmanned Aerial Vehicle Data and the satellite data progress from the acquisition of satellite remote sensing film source
Match, obtain pretreatment satellite photo.
First, the multi-wavelength data and panchromatic wave-band data of remote sensing image data (satellite data) are merged, makes it
It is distinguished as 0.5 meter.Utilize country 1:2000 topographic map carries out multinomial coefficient geometric correction.Specifically：To the original of purchase
High-definition remote sensing data product carries out geometric correction, finds on 5~10 pairs of topographic maps and remote sensing image data on same position
Control point (being usually the high road junction of resolution, bridge or mark house etc.), it is several to the respective of control point by this
Coordinate (XY values) utilizes multinomial coefficient algorithm, equation coefficients when calculating error is minimum, by the formula, to remote sensing image
Upper all pixels point is converted, and completes the geometric correction of whole figure.
Pass through geometric correction, it can be ensured that survey region and check post matching are identical in satellite data and unmanned plane photo
Region.Then, unmanned plane photo GPS information and high spatial resolution remote sensing data (VHR data) are imported by ArcGIS softwares
Matching, will be in photo using the Geotaggedphotos to point tools of the Data Management in ArcGIS softwares
GPS information based on EXIF switchs to vector dot format, and passes through identify query facilities and can trace to the source original photo vegetation information,
Accordingly for vector point data addition vegetation and plant species information, species attribute name of the information and finally in eCognitionDeveloper
Software classification class of procedure title will be consistent.
6th step：Patch properties information in extraction pretreatment satellite photo.
Utilize segmentation (Segmentation), the classification in eCognitionDeveloper softwares
(Classification) function, the spectrum of invasive species and other species patches where extraction unmanned plane sampling point, shape, position
Confidence ceases.
In segmentation and classification algorithm, multi-scale division first is carried out to VHR remote sensing images (high spatial resolution remote sensing data)
(Multiresolution segmentation) is determined for compliance with the best scale (Scale) of most of Plant patch feature；
Under this best scale, by chessboard split plot design (Chessboard segmentation), obtain identical with sample point and carry class
The smallest partition object of other information, sampling point patch used in analysis as in later-mentioned step.Utilize sorting algorithm (Assign
Class by thematic layer) classified by the attribute (Class_name) of sample point to sampling point figure spot；Continue profit
The sample patch being overlapped with sample point object imparting and sample are ordered the same classification with sorting algorithm (Assign class) (to lead to
Cross addition attribute restrictive condition：Border to class attributes in Class-related features), institute is obtained at this time
Homogeneous patch where having the sampling point of known vegetation and plant species.
By the Patch properties selection function in eCognitionDeveloper softwares, select patch spectrum, space special
Sign：The index characteristic instruction for creating and needing is added at Feature View interfaces, such as：NDVI, layer mean/standard
Deviation values, geometry-extent-area/length/width, position-distance to the
Scene border, geometry-shape-asymmetry/border index/compactness/main
Direction/shape index, texture-layer value texture based on sub-object,
Texture-shape value texture based on sub-object, it is empty to obtain the correlation of known sampling point patch
Between characteristic value.
7th step：Optimal separability characteristic index is obtained according to Patch properties information sifting.
Utilize use output object statistic algorithm (the Export object in eCognitionDeveloper softwares
Statistics), set level, class filter, Features, export each attribute information of each Plant patch (spectrum,
Shape, position) statistical data, then counted the validity feature analyzed contrast to different classes of classification again.
Multivariate analysis of variance (analysis is carried out by SPSS softwares>General linear model>Multivariable), to different type plant
The sample average otherness of patch attribute carries out significance test, filters out optimal separability characteristic index, and determine these
Index is used to distinguish the threshold range of every a kind of vegetation.
8th step：The precision of satellite data is evaluated according to optimal separability characteristic index.
The imaged object that first the 4th step is extracted, i.e., categorized sample patch convert the subsample of constituent class
(Classified image objects to samples)；Again based on obtained optimal separability characteristic index,
Sorting algorithm rule set is established in eCognition Developer softwares, whole picture VHR remote sensing images are carried out according to set of algorithms
Supervised classification in the present embodiment, selects nearest classification (Nearest neighbor classification) to VHR
Remotely-sensed data is classified, and is passed through the error moments tactical deployment of troops (Error matrix based on TTA mask) progress precision and commented
Valency.
Error matrix (error matrix) method is the method for a kind of precision evaluation, applies to carry out using sample characteristics
Accuracy Assessment after supervised classification.It is one and is used for the pixel number for representing to be divided into a certain classification and ground check as such
Not Shuo comparator array.In general, the row in array represent reference data, row represents the classification number classified by remotely-sensed data
According to.There are pixel number and percentage to represent two kinds.It is missed from error matrix overall classification accuracy, Kappa coefficients, misclassification error, leakage point
Poor, every a kind of cartographic accuracy and user's precision determine the precision and reliability of classification by these accuracy values.
The present invention is described in detail above in association with attached drawing embodiment, those skilled in the art can be according to upper
It states and bright many variations example is made to the present invention.Thus, certain details in embodiment should not form limitation of the invention, this
Invention will be using the range that the appended claims define as protection scope of the present invention.
Claims (8)
1. a kind of estuarine wetland denizen monitoring method, including step：
S1：A unmanned aerial vehicle platform and a satellite remote sensing film source are selected according to monitoring requirements；
S2：The setting at least course line of a unmanned plane and the corresponding flight ginseng to unmanned plane setting in a Target monitoring area
Number；
S3：One Unmanned Aerial Vehicle Data acquisition tasks are set according to the climatic data of Target monitoring area；
S4：The Unmanned Aerial Vehicle Data acquisition tasks are performed, obtain Unmanned Aerial Vehicle Data；
S5：The geography information of the Unmanned Aerial Vehicle Data is matched with the satellite data obtained from the satellite remote sensing film source,
Obtain pretreatment satellite photo；
S6：Multiple patches are formed, and the patch for forming the patch is special in the pretreatment satellite photo according to floristics
Reference ceases；
S7：Optimal separability characteristic index is obtained according to the Patch properties information sifting；
S8：Classified according to the optimal separability characteristic index to the satellite data and precision evaluation is carried out to result.
2. estuarine wetland denizen monitoring method according to claim 1, which is characterized in that described in the S1 steps
Satellite remote sensing film source include Pleiades remote sensing high resolution images source, QuickBird remote sensing high resolution image sources and
IKONOS remote sensing high resolution images source.
3. estuarine wetland denizen monitoring method according to claim 2, which is characterized in that described in the S2 steps
Flight parameter includes course line, shooting height, shooting angle, flying speed and shooting interval；According to the satellite remote sensing film source
The satellite data sets the course line, shooting height and shooting angle；The flying speed is set according to the course line；According to
The flying speed sets the shooting interval, and the unmanned aerial vehicle has shadow according to the adjacent photo that the shooting interval is shot
As overlapping region；Obtain the intertidal zone cross-sectional distance d of the Target monitoring area.
4. estuarine wetland denizen monitoring method according to claim 3, which is characterized in that the S3 steps are further wrapped
Include step：
S31：Obtain the tide table of the Target monitoring area；
S32：Elevation of high water, elevation of low water and the time of tide of the Target monitoring area spring season is obtained according to the tide table；
S33：According to the farthest flying distance D of the unmanned plane and intertidal zone cross-sectional distance d, the unmanned plane is determined
Shooting performs the period, forms the Unmanned Aerial Vehicle Data acquisition tasks；
As D ＞ d, the shooting execution period of the unmanned plane was located at before the time of tide of the elevation of high water, the nothing
The duration that man-machine shooting performs the period is more than or equal to the two-way time that the unmanned plane is shot along the course line；
As D≤d, staff need to carry unmanned plane be moved to apart from waterside from water distance from d ' less than it is described it is farthest flight away from
Entered back into during from D the unmanned plane shooting perform the period, the unmanned plane shooting perform the period duration be more than etc.
The sum of the two-way time shot in the unmanned plane along the course line and the two-way time of the staff.
5. estuarine wetland denizen monitoring method according to claim 4, which is characterized in that the S5 steps are further wrapped
Include step：
S51：The satellite data is carried out to the pretreatment of Band fusion and geometric correction, obtains pretreatment remotely-sensed data；
S52：The Unmanned Aerial Vehicle Data is imported by GIS software；
S53：By the Unmanned Aerial Vehicle Data vector quantization, the Unmanned Aerial Vehicle Data after vector quantization includes picture data, the first GPS believes
Breath and floristics information, each first vector point and first GPS information and the floristics of the picture data are believed
Breath association；
S54：It is described pretreatment remotely-sensed data include satellite image data, the second GPS information, unknown floristics spectral information and
Spatial information；Using first GPS information and the consistency of second GPS information, by each of the satellite image data
Second vector point and associated each information match described in the first vector point of the picture data so that the unknown plant
Species spectral information and the floristics information match.
6. estuarine wetland denizen monitoring method according to claim 5, which is characterized in that in the S6 steps, utilize
The dividing function of eCognitionDeveloper softwares obtains same plant model shared in the pretreatment satellite photo
It encloses, multiple patches is formed in the pretreatment satellite photo according to floristics；Utilize eCognitionDeveloper softwares
Classification feature be associated with the corresponding floristics information for the patch assignment；Form Patch properties information, the patch
Characteristic information includes：Patch spectral information, patch shape information and plaque location information.
7. estuarine wetland denizen monitoring method according to claim 6, which is characterized in that the S7 steps are further wrapped
Include step：
Using the output function in the eCognitionDeveloper softwares, the statistical number of the Patch properties information is exported
According to；
By statistical data described in multivariate analysis of variance, filter out the optimal separability characteristic index and determine per a kind of vegetation
Threshold range.
8. estuarine wetland denizen monitoring method according to claim 7, which is characterized in that the S8 steps are further wrapped
Include step：
Sorting algorithm rule is established by the eCognitionDeveloper softwares based on the optimal separability characteristic index
Collection；
Supervised classification using the set of algorithms in the eCognitionDeveloper softwares divides the satellite data
Class obtains a classification pixel set；
By the classification pixel set compared with a reference image metaset cooperation, the precision of the satellite data is evaluated.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711416470.6A CN108195767B (en) | 2017-12-25 | 2017-12-25 | Estuary wetland foreign species monitoring method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711416470.6A CN108195767B (en) | 2017-12-25 | 2017-12-25 | Estuary wetland foreign species monitoring method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108195767A true CN108195767A (en) | 2018-06-22 |
| CN108195767B CN108195767B (en) | 2020-07-31 |
Family
ID=62583794
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711416470.6A Active CN108195767B (en) | 2017-12-25 | 2017-12-25 | Estuary wetland foreign species monitoring method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108195767B (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110390255A (en) * | 2019-05-29 | 2019-10-29 | 中国铁路设计集团有限公司 | High-speed rail environmental change monitoring method based on various dimensions feature extraction |
| 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 |
| CN111553244A (en) * | 2020-04-24 | 2020-08-18 | 中国电建集团成都勘测设计研究院有限公司 | Water and soil conservation monitoring method based on automatic positioning and orienting technology |
| CN111767643A (en) * | 2020-06-03 | 2020-10-13 | 中国水产科学研究院东海水产研究所 | Method for determining distribution range of biological subject environmental factors |
| 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 |
| CN114414701A (en) * | 2022-01-28 | 2022-04-29 | 辽宁中医药大学 | Preparation method and content determination method of myrobalamin acid in myrobalan meat |
| CN115810155A (en) * | 2023-01-18 | 2023-03-17 | 中关村睿宸卫星创新应用研究院 | Tidal wetland classification method |
| CN115834845A (en) * | 2023-02-24 | 2023-03-21 | 山东慧创信息科技有限公司 | Unmanned aerial vehicle land data identification system based on thing networking |
| CN116625354A (en) * | 2023-07-21 | 2023-08-22 | 山东省国土测绘院 | High-precision topographic map generation method and system based on multi-source mapping data |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103528570A (en) * | 2013-10-23 | 2014-01-22 | 环境保护部卫星环境应用中心 | Spatial distribution acquiring method and system for spartina alterniflora in northern sea area of Guangxi |
| CN104463470A (en) * | 2014-12-10 | 2015-03-25 | 中国水产科学研究院东海水产研究所 | Intertidal zone habitat survey method based on unmanned aerial vehicle aerial photography |
| CN104915674A (en) * | 2014-10-24 | 2015-09-16 | 北京师范大学 | Landsat8 and MODIS fusion-construction high space-time resolution data identification autumn grain crop method |
| CN106915462A (en) * | 2017-02-14 | 2017-07-04 | 福建兴宇信息科技有限公司 | Forestry pests ＆ diseases intelligent identifying system based on multi-source image information |
| CN107389036A (en) * | 2017-08-02 | 2017-11-24 | 珠江水利委员会珠江水利科学研究院 | A kind of large spatial scale vegetation coverage computational methods of combination unmanned plane image |
| CN107478611A (en) * | 2017-08-02 | 2017-12-15 | 珠江水利委员会珠江水利科学研究院 | A kind of method for calculating the exposed rate of rock |
-
2017
- 2017-12-25 CN CN201711416470.6A patent/CN108195767B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103528570A (en) * | 2013-10-23 | 2014-01-22 | 环境保护部卫星环境应用中心 | Spatial distribution acquiring method and system for spartina alterniflora in northern sea area of Guangxi |
| CN104915674A (en) * | 2014-10-24 | 2015-09-16 | 北京师范大学 | Landsat8 and MODIS fusion-construction high space-time resolution data identification autumn grain crop method |
| CN104463470A (en) * | 2014-12-10 | 2015-03-25 | 中国水产科学研究院东海水产研究所 | Intertidal zone habitat survey method based on unmanned aerial vehicle aerial photography |
| CN106915462A (en) * | 2017-02-14 | 2017-07-04 | 福建兴宇信息科技有限公司 | Forestry pests ＆ diseases intelligent identifying system based on multi-source image information |
| CN107389036A (en) * | 2017-08-02 | 2017-11-24 | 珠江水利委员会珠江水利科学研究院 | A kind of large spatial scale vegetation coverage computational methods of combination unmanned plane image |
| CN107478611A (en) * | 2017-08-02 | 2017-12-15 | 珠江水利委员会珠江水利科学研究院 | A kind of method for calculating the exposed rate of rock |
Non-Patent Citations (6)
| Title |
|---|
| LU BING等: "Species classification using Unmanned Aerial Vehicle (UAV)-acquired high spatial resolution imagery in a heterogeneous grassland", 《ISPRS JOURNAL OF PHOTOGRAMMETRY AND REMOTE SENSING》 * |
| MARIA CALVINO-CANCELA等: "Alien Plant Monitoring with Ultralight Airborne Imaging Spectroscopy", 《PLOS ONE》 * |
| WAN HUAWEI等: "Monitoring the Invasion of Spartina alterniflora Using Very High Resolution Unmanned Aerial Vehicle Imagery in Beihai, Guangxi (China)", 《THE SCIENTIFIC WORLD JOURNAL》 * |
| WANG ANQI等: "Monitoring the Invasion of Spartina alterniflora from 1993 to 2014 with Landsat TM and SPOT 6 Satellite Data in Yueqing Bay, China", 《PLOS ONE》 * |
| ZHOU ZAIMING等: "Estimating Spartina alterniflora fractional vegetation cover and aboveground biomass in a coastal wetland using SPOT6 satellite and UAV data", 《AQUATIC BOTANY》 * |
| 王利民等: "基于无人机影像的农情遥感监测应用", 《农业工程学报》 * |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110390255A (en) * | 2019-05-29 | 2019-10-29 | 中国铁路设计集团有限公司 | High-speed rail environmental change monitoring method based on various dimensions feature extraction |
| 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 |
| CN111553244A (en) * | 2020-04-24 | 2020-08-18 | 中国电建集团成都勘测设计研究院有限公司 | Water and soil conservation monitoring method based on automatic positioning and orienting technology |
| CN111767643A (en) * | 2020-06-03 | 2020-10-13 | 中国水产科学研究院东海水产研究所 | Method for determining distribution range of biological subject environmental factors |
| 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 |
| CN114414701A (en) * | 2022-01-28 | 2022-04-29 | 辽宁中医药大学 | Preparation method and content determination method of myrobalamin acid in myrobalan meat |
| CN115810155A (en) * | 2023-01-18 | 2023-03-17 | 中关村睿宸卫星创新应用研究院 | Tidal wetland classification method |
| CN115810155B (en) * | 2023-01-18 | 2023-04-18 | 中关村睿宸卫星创新应用研究院 | Tidal wetland classification method |
| CN115834845A (en) * | 2023-02-24 | 2023-03-21 | 山东慧创信息科技有限公司 | Unmanned aerial vehicle land data identification system based on thing networking |
| CN116625354A (en) * | 2023-07-21 | 2023-08-22 | 山东省国土测绘院 | High-precision topographic map generation method and system based on multi-source mapping data |
| CN116625354B (en) * | 2023-07-21 | 2023-09-29 | 山东省国土测绘院 | High-precision topographic map generation method and system based on multi-source mapping data |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108195767B (en) | 2020-07-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108195767A (en) | Estuarine wetland denizen monitoring method | |
| Torres-Sánchez et al. | Mapping the 3D structure of almond trees using UAV acquired photogrammetric point clouds and object-based image analysis | |
| CN105740759B (en) | Semilate rice information decision tree classification approach based on feature extraction in multi-temporal data | |
| CN107527014A (en) | Crops planting area RS statistics scheme of sample survey design method at county level | |
| CN112287892A (en) | Arbor biomass measurement and calculation method based on unmanned aerial vehicle hyperspectral and machine learning algorithm | |
| Ainiwaer et al. | Deep learning-based rapid recognition of oasis-desert ecotone plant communities using UAV low-altitude remote-sensing data | |
| CN115641412A (en) | Hyperspectral data-based three-dimensional semantic map generation method | |
| Lass et al. | Detecting the locations of Brazilian pepper trees in the Everglades with a hyperspectral sensor | |
| Zou et al. | The fusion of satellite and unmanned aerial vehicle (UAV) imagery for improving classification performance | |
| Cui et al. | Optimal spatial resolution of remote-sensing imagery for monitoring cantaloupe greenhouses | |
| Frimpong | Application of Remote Sensing and GIS for Forest Cover Change Detection.(A Case Study of Owabi Catchment in Kumasi, Ghana) | |
| Bruce | Object oriented classification: case studies using different image types with different spatial resolutions | |
| Frimpong | Application of remote sensing and GIS for forest cover change detection | |
| Deng et al. | Tree crown recognition algorithm on high spatial resolution remote sensing imagery | |
| Li et al. | Application of High-Resolution Remote Sensing Image for Individual Tree Identification of Pinus sylvestris and Pinus tabulaeformis | |
| Chib | Drones for conservation: integrating UAVs with field methods to classify satellite imagery to map plant communities–A case study of Drentsche AA, The Netherlands | |
| Barlow | A comparative study of raster and vector based approaches in vegetation mapping on Five Islands off the coast of Port Kembla. | |
| Oliech | Geo-spatial assessment of forest health status using UAS technology in Anopoli, Crete | |
| Mattilio et al. | Multispectral unmanned aerial system remote sensing to detect Dalmatian toadflax in a mixed sagebrush steppe on the South Fork of the Shoshone River | |
| Jia et al. | Long Image Time Series for Crop Extraction Based on the Automatically Generated Samples Algorithm | |
| Deng et al. | Automated extracting tree crown from QuickBird Stand Image | |
| Xiong et al. | Detecting and Mapping Individual Fruit Trees in Complex Natural Environments via UAV Remote Sensing and Optimized YOLOv5 | |
| Machala et al. | Generating land-cover maps from remotely sensed data: manual vectorization versus object-oriented automation. | |
| Yasen | UAV-Based Sorghum Growth Monitoring Using LiDar, Photogrammetry, and Machine Learning | |
| LOONG | OIL PALM DETECTION AND DELINEATION USING LOCAL MAXIMA, TEMPLATE MATCHING AND SEEDED REGION GROWING |
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 |