CN107909607B - A kind of year regional vegetation coverage computational methods - Google Patents
A kind of year regional vegetation coverage computational methods Download PDFInfo
- Publication number
- CN107909607B CN107909607B CN201711310827.2A CN201711310827A CN107909607B CN 107909607 B CN107909607 B CN 107909607B CN 201711310827 A CN201711310827 A CN 201711310827A CN 107909607 B CN107909607 B CN 107909607B
- Authority
- CN
- China
- Prior art keywords
- value
- vegetation
- ndvi
- index information
- remote sensing
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/30—Determination of transform parameters for the alignment of images, i.e. image registration
- G06T7/33—Determination of transform parameters for the alignment of images, i.e. image registration using feature-based methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/188—Vegetation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10032—Satellite or aerial image; Remote sensing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30181—Earth observation
- G06T2207/30188—Vegetation; Agriculture
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Geometry (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Multimedia (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a kind of annual regional vegetation coverage computational methods, including：Accurate radiation calibration, atmospheric correction and geometrical registration are carried out to the same annual all remote sensing images in region, obtain remote sensing image result set of graphs；Cloud layer index, shadow index, background index based on structure and vegetation index, extract corresponding thematic information；Cloud algorithm is carried out respectively to different phase remote sensing images and shadow algorithm is gone to handle；Space combination is carried out to Mono temporal vegetation index information；Multidate information in year is merged, computation year regional vegetation index information spatial distribution map；Finally, regional vegetation computation year coverage information space distribution map.Influence this invention removes factors such as weather, landform, phase, vegetation pattern, environment to vegetation coverage, solve the problems, such as that annual regional vegetation coverage calculates, improve accuracy and the reliability of the calculating of regional scale vegetation coverage, operating process is simple, flexible, is easy to promote and apply in regional scale.
Description
Technical field
The present invention relates to a kind of annual regional vegetation coverage computational methods, belong to ecological construction, ecological environmental protection, calamity
The application fields such as evil monitoring, ecological functions assessment.
Background technology
Vegetation coverage is normally defined the planimetric area of vegetation (leaf, stem, branch) on ground and accounts for the Statistical Area gross area
Percentage, be the important indicator for one of leading indicator and local ecosystem environmental change for weighing surface vegetation situation,
In occupation of critical role in atmospheric thermodynamics, pedosphere, hydrosphere and biosphere.Vegetation coverage is considered as carrying out soil desert
Most effective index in the land deteriorations such as change, stony desertification, salination assessment；It reflects shadow of the ecosystem to the soil erosion
It rings, is the active factor for controlling the soil erosion, be important indicator and the hydrology life for carrying out soil erosion assessment and comprehensive treatment
A significant variable in states model, and then there is weight in local area ecological functional assessment and resource environment comprehensive bearing capacity assessment
Want meaning；Meanwhile vegetation coverage and vegetation evapotranspire it is close contact, and it is energy balance and water balance that vegetation, which is evapotranspired,
Important component, be soil-vegetation-atmosphere system water and heat transmission in a process, the monitoring needs root to evapotranspire
Different methods is used according to different vegetation coverages, and vegetation coverage is controlling elements important as one and exists；In addition,
Vegetation coverage is also a sensitive factor in global change research due, is many regions or even global climate model, carbon process
An essential variable in assessment models can be variation of ecology and environment, regional vegetation produces force evaluating, land cover pattern variation is supervised
It surveys and useful information is provided.Therefore, it is each application neck to carry out the calculating of regional vegetation coverage and its raising of computational accuracy
Domain development is required, is utilized to ecological construction, environmental protection, the hydrology, agricultural, land resources, disaster monitoring and the whole world become
Change research etc. to be all of great significance.
Currently, the computational methods of vegetation coverage can substantially be divided into two kinds：It is another one is traditional ground measurement method
Kind is remote sensing monitoring method.
Although traditional ground measurement method measurement accuracy is higher, this method is time-consuming, laborious, in addition vegetation coverage itself
With significant Spatio-temporal Distribution, this method is not suitable for regional scale research, and the research for being only applicable to smaller spatial dimension is answered
In.
The appearance of remote sensing technology is to monitor region vegetation coverage, or even global vegetation coverage variation provides possibility.
The method of existing remote sensing monitoring vegetation coverage has very much, includes mainly：Empirical model method, Decomposition of Mixed Pixels method, physics mould
Type method, spectrum gradient method.Empirical model method is to be based on measured data, establishes corresponding statistical model or empirical regression model, into
And calculate vegetation coverage；Although this method is simple, is easy to calculate, need a large amount of measured data to establish correlation model,
Meanwhile needing to establish different empirical models for different vegetation types, easily by restrictions such as time, regional condition, vegetation patterns,
There is larger uncertainty and limitation in big regional study, be not easy to promote and apply.Pixel Unmixing Models method is also referred to as
The remote sensing information in the pixel that multiple components are constituted is decomposed, pixel is established that is, under certain assumed condition for Sub-pixel decompound method
Decomposition model, to obtain vegetation coverage.Pixel Unmixing Models mainly linear model, probabilistic model and fuzzy analysis
Model etc., wherein linear unmixed model is most widely used.Linear unmixed model assume reach sensor photon only with one
Component acts on, and different component is mutual indepedent, is corresponding vegetation by ratio of the Numerical Methods Solve each component in pixel
Coverage, precision is heavily dependent on the Rational choice of each end member, therefore it is not suitable for the remote sensing of low spatial resolution
Data application.Common linear unmixed model has pixel dichotomy, artificial neural network method, line spectrum analytic approach, multiterminal member
Spectral mixing analysis etc., such method require the higher remotely-sensed data of spatial resolution.Physical model method is by studying light
Interaction with vegetation theoretically has wide to establish the model of vegetation spectral information and vegetation coverage physical relation
Application, but this class model needs a large amount of associated parameter data, and existing satellite remote sensing date is in the feelings for ensureing model accuracy
Condition, it is difficult to meet parameter request, and correlation model calculating is also more complicated, difficulty or ease are promoted in regional scale range.Spectrum ladder
Degree method is to be proposed on the basis of analyzing vegetation and soil reflects, and assume that spectral reflectance is with wavelength line within the scope of limited wavelength band
Property variation, do not account for variation of the vegetation soil area with wavelength, model error is larger.
With the progress of remote sensing technology, the spatial resolution of remote sensing image has no longer been the accurate limit for obtaining vegetation coverage
The revisiting period of the factor processed, remote sensing also greatly shortens.At the same time, because of factors such as weather, landform, sensor itself, image is made an uproar
Sound (such as cloud layer, shade, background information) also increases therewith, how to be effectively reduced the influence of noise, while farthest
It has been that the technology for being badly in need of solving instantly is asked using the useful information in remote sensing image to improve the accuracy of vegetation coverage calculating
Topic.In addition, existing calculate in research application vegetation coverage, a relatively good phase of this area's vegetation growth is generally selected
Data represent the vegetation coverage of this area.And a regional vegetation cover type is often varied in reality,
Each type of vegetation growth season is all different；Even same type of vegetation, because of edaphic condition, moisture condition, kind
Plant the difference that the differences such as time also result in growth situation.Therefore, it is necessary to consider in regional extent different vegetation types and
The influence of growing environment, phase data when making full use of multiple, to improve accuracy, the reliability of the calculating of regional vegetation coverage.
Existing application needs not only to the vegetation coverage information of true representations this area, especially with the expansion in research range region
Greatly, the influence that the above problem accurately calculates vegetation coverage can be more and more significant.
Invention content
In view of the deficienciess of the prior art, the present invention proposes a kind of annual regional vegetation coverage computational methods, it should
Method has fully considered the interference of the noises such as cloud layer, shade, background, can excavate and utilize to the maximum extent remote sensing image information,
The accuracy that existing vegetation coverage calculates is improved, operating process is simple, flexible, convenient to be promoted on larger regional scale
Using.
To solve the above problems, the technical solution used in the present invention is：
A kind of year regional vegetation coverage computational methods, include the following steps：
S1 carries out accurate radiation calibration and atmospheric radiation to same annual all remote sensing images of vegetation growing season inner region
Correction, obtains the remote sensing image result figure that different phases represent atural object real reflectance；
S2, for remote sensing image result figure obtained by step S1, on the basis of any phase remote sensing image result figure, to other
Remote sensing image result figure carries out accurate geometrical registration, forms a remote sensing image result to match on geospatial coordinates
Set of graphs；
S3, for each phase remote sensing image result figure in remote sensing image result set of graphs obtained by step S2 carry out respectively with
Lower processing：
S3-1 calculates separately cloud layer index, vegetation index and shadow index for the phase remote sensing image result figure, obtains
Corresponding index information spatial distribution map；
S3-2 calculates background index for the phase remote sensing image result figure, and carries out binary conversion treatment, obtains a shadow
As background information spatial distribution map；
S3-3 carries out cloud algorithm process, the algorithm packet for vegetation index information space distribution map obtained by step S3-1
Include following sub-step：
S3-3-1 carries out the phase remote sensing image result figure enhancing algorithm process of image, and by enhancing algorithm
The pixel sample in random acquisition cloud layer area is sample area in remote sensing image result figure that treated；
Cloud layer index information spatial distribution map obtained by the sample area and step S3-1 is carried out space overlapping, system by S3-3-2
Meter calculates sample area and corresponds to the cumulative frequency of cloud layer index information numerical value from small to large；
According to the cumulative frequency threshold range is arranged, and make two and cloud layer index information spatial distribution in S3-3-3
The two-value hum pattern that figure matches；
S3-3-4, by vegetation index information space distribution map and step obtained by two two-value hum patterns of making and step S3-1
Image background information spatial distribution map carries out spatial operation obtained by rapid S3-2, you can the vegetation index after the cloud layer noise that is eliminated
Information result figure；
S3-4 carries out shadow algorithm and handles for vegetation index information space distribution map obtained by step S3-1, the algorithm
Including following sub-step：
S3-4-1 carries out the phase remote sensing image result figure enhancing algorithm process of image, and by enhancing algorithm
The pixel sample in random acquisition shadow region is sample area in remote sensing image result figure that treated；
Shadow index information space distribution map obtained by the sample area and step S3-1 is carried out space overlapping, system by S3-4-2
Meter calculates that sample area corresponds to shadow index information value cumulative frequency from small to large and sample area corresponds to shadow index Information Number
Value；
S3-4-3 corresponds to shadow index information value according to the cumulative frequency and sample area, a threshold range is arranged, and make
Make two two-value hum patterns to match with shadow index information space distribution map；
S3-4-4, by vegetation index information data space distribution map obtained by two two-value hum patterns of making and step S3-1
Spatial operation is carried out with image background information spatial distribution map obtained by step S3-2, you can the vegetation after the shade noise that is eliminated
Index information result figure；
S3-5, based on the image background information spatial distribution map and step S3-3 and step S3-4 made by step S3-2
Made two-value hum pattern makes two new two-value hum patterns, and then utilizes new two-value hum pattern, to step S3-3 and
Vegetation index information result figure obtained by step S3-4 carries out the space combination of Mono temporal information, and the final vegetation for generating the phase refers to
Number information result figure；
S4, based on the final vegetation index information result figure of each phase obtained by step S3, by year all phases in region
Final vegetation index information result figure merged, formed an image data for including multiple wave bands；
S5, on the bases step S4, to covering the final vegetation index information result figure of same area, using maximum value
Synthetic method carries out maximum value synthesis processing, obtains year regional vegetation index information spatial distribution map；
S6 carries out numerical operation processing to regional vegetation index information spatial distribution map obtained by step S5, eliminates cloud layer, the moon
The influence of shadow and background information to statistic analysis result；Accurately and reliably data source is provided to be for statistical analysis in next step；
S7 is based on step S6 acquired results, calculates vegetation coverage to get the accurate vegetation coverage in the year region
Spatial distribution map.
Further, the cloud layer index CCI built in the step S3-1, calculation formula are：CCI=(ρR+ρG+ρB)*
ρNIR, ρRFor the spectral reflectance values of red wave band, ρGFor the spectral reflectance values of green wave band, ρBFor the spectral reflectivity of blue wave band
Value, ρNIRFor remote sensing image near infrared band spectral reflectance values.
Further, vegetation index is normalized differential vegetation index NDVI in the step S3-1, which is：ρNIRFor remote sensing image near infrared band spectral reflectance values, ρRFor the spectral reflectivity of red wave band
Value.
Further, shadow index YYI calculation formula described in the step S3-1 are：Its
Middle ρNIRFor remote sensing image near infrared band spectral reflectance values, ρSWIRFor the spectral reflectance values of short infrared wave band, NDVI is
Normalized differential vegetation index.
Further, the background index BGI built in the step S3-2, calculation formula are：Wherein ρNIRFor remote sensing image near infrared band spectral reflectance values, ρRSpectrum for red wave band is anti-
Radiance rate value, ρGFor the spectral reflectance values of green wave band.
Further, binary conversion treatment refers to by background information pixel assignment in background index BGI in the step S3-2
It is -99, other pixels are assigned a value of 0, obtain an image background information spatial distribution map BGII.
Further, in the step S3-3-1 enhancing algorithm process of image refer to 2% linear stretch enhancing；At random
Collecting sample pixel refers to the sampling window random acquisition cloud layer information pixel using 2 × 2, to avoid mistake from choosing non-cloud layer area picture
Member and cause error, it is desirable that sample pixel quantity be more than 200.
Further, the setting of threshold range refers to cloud layer when cumulative frequency is reached 2% in the step S3-3-3
Lower limit of the index information numerical value as threshold value, the upper threshold limit threshold value are infinity.
Further, two two-value hum patterns are CCIB1 and CCIB2 in the step S3-3-3, and CCIB1 is to be directed to cloud layer
Numerical value in threshold range is assigned a value of 0 by index CCI, and other values are assigned a value of 1, calculate and obtain；CCIB2 is to be directed to cloud layer index
Numerical value in threshold range is assigned a value of -99 by CCI, and other values are assigned a value of 0, calculate and obtain.
Further, the calculation formula of spatial operation is in the step S3-3-4：NDVICCI=NDVI*CCIB1+
CCIB2+BGII, wherein NDVICCIFor the vegetation index information result figure for eliminating after cloud layer noise, BGII is image background information
Spatial distribution map, CCIB1 and CCIB2 are that step S3-3-3 calculates gained two-value hum pattern.
Further, the enhancing algorithm process of image described in the step S3-4-1 refers to 2% linear stretch enhancing；
Random acquisition sample pixel refers to the sampling window random acquisition image shades information pixel using 2 × 2, non-to avoid mistake from choosing
Cloud layer area pixel and cause error, it is desirable that sample pixel quantity be more than 200.
Further, the setting of threshold range described in the step S3-4-3 refers to when cumulative frequency is reached 95%
The upper limit of the shadow index information value as threshold value, sample area correspond to the minimum value of shadow index information value as threshold range
Lower limit.
Further, two two-value hum patterns described in the step S3-4-3 are YYIB1 and YYIB2, and YYIB1 is to be directed to
Numerical value in threshold range is assigned a value of 0 by shadow index YYI, and other values are assigned a value of 1, calculate and obtain；YYIB2 is to be directed to shade
Numerical value in threshold range is assigned a value of -99 by index YYI, and other values are assigned a value of 0, calculate and obtain.
Further, the calculation formula of spatial operation is in the step S3-4-4：NDVIYYI=NDVI*YYIB1+
YYIB2+BGII, wherein NDVIYYIFor the vegetation index information result figure for eliminating after shadow information, BGII is that step S3-2 is calculated
Gained image background information spatial distribution map, YYIB1 and YYIB2 are that step S3-4-3 calculates gained two-value hum pattern.
Further, two new two-value hum patterns are NDVI in the step S3-5B1And NDVIB2, calculation formula point
It is not as follows：
NDVIB2=(NDVIB1+99)*(NDVIB1-1)
Wherein BGII is that step S3-2 calculates gained background information spatial distribution map, and CCIB1 is that step S3-3 calculates gained
Two-value hum pattern, YYIB1 are that step S3-4 calculates gained two-value hum pattern.
Further, space combination method, calculation formula described in the step S3-5 are as follows：
Wherein NDVICCIFor the vegetation index information result figure for eliminating after cloud layer noise obtained by step S3-3, NDVIYYIFor step
The vegetation index information result figure after shadow information, NDVI are eliminated obtained by rapid S3-4B1And NDVIB2It is new for two in step S3-5
Two-value hum pattern.
Further, fusion described in step S4 refers to synthesizing to obtain by wave band by multiple Mono temporal vegetation index information
One image spatially to match.
Further, maximum value synthetic method refers to being directed to the same pixel in the step S5, and the pixel is in different phases
Different numerical value is had, the maximum value in these values, as maximum value is taken to synthesize.
Further, the processing of numerical operation described in step S6 refers to by regional vegetation index information space obtained by step S5
The area assignment that numerical value is -99 in distribution map is NAN, and accurately and reliably data source is provided to be for statistical analysis in next step.
Further, in the step S7 calculate vegetation coverage can using any type based on vegetation index information come
Estimate the method or model of vegetation coverage.
Further, the method for any type estimation vegetation coverage or model include normal in Decomposition of Mixed Pixels method
Pixel dichotomy, neural network model.
Further, the pixel dichotomy calculation formula is as follows：
Wherein VFC is the vegetation coverage in pixel, and NDVI is the normalized differential vegetation index on the pixel, NDVIsoilFor
NDVI values without vegetative coverage pixel, NDVIvegFor completely by the pixel NDVI values of vegetative coverage；In the present invention, with 0.5%
The upper lower threshold value of confidence level interception area NDVI respectively represents NDVIvegAnd NDVIsoil, and then zoning vegetation coverage
VFC。
The beneficial effects of the present invention are：
1) it is an advantage of the invention that from analysis weather, landform, when the conditions such as phase character regional vegetation coverage is calculated
Influence is started with, and is constructed the model algorithms such as cloud layer index, shadow index, background index, is farthest eliminated environmental background
Influence of the noise to vegetation coverage result of calculation solves the problems, such as that annual regional vegetation coverage calculates, improves region
The accuracy and reliability that scale vegetation coverage calculates, operating process is simple, flexible, is easy to push away in Application in regional scale
Extensively.
2) method, the present invention take full advantage of the high time resolution, multidate of remotely-sensed data now compared with prior art
Feature considers vegetation pattern in region, the difference of growing environment, Growing season, it is proposed that a kind of year vegetation coverage meter
Calculation method improves vegetation coverage as measurement surface vegetation situation, the standard of local ecosystem environmental change leading indicator
Exactness, reliability and confidence level, and as ecology, the accuracy of the hydrology, climate model necessity input parameter, have relatively strong
Universality and adaptability, easily promote and apply.
Description of the drawings
Fig. 1 is the flow diagram of the present invention；
Fig. 2 is the remote sensing image of four different phases；
Fig. 3 is cloud and goes the spatial operation principle schematic of shade；
Fig. 4 is vegetation index information result figure of four phases through past cloud and after going Shadows Processing；
Fig. 5 is annual regional vegetation index information spatial distribution map；
Fig. 6 is the accurate vegetation coverage spatial distribution map in annual region.
Specific implementation mode
In order to make the purpose , technical scheme and advantage of the present invention be clearer, with reference to the accompanying drawings and embodiments, right
The present invention is further elaborated.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, and
It is not used in the restriction present invention.
The present embodiment by certain region with year, four different phases remote sensing image for illustrate.
Referring to Fig.1, Fig. 1 is a kind of flow diagram of annual regional vegetation coverage computational methods of the present invention, including with
Lower step：
S1 collects all remote sensing images in the same annual Growing season in one's respective area, carries out accurate radiation calibration to it respectively
And Atmospheric radiation correction, obtain the remote sensing image result figure that can represent atural object real reflectance.If Fig. 2 is the phase of one's respective area four
Remote sensing image, the first width figure (the first row left figure) is the remote sensing image in June in Fig. 2, the second width figure (first in Fig. 2
Row right figure) be July remote sensing image, third width figure (the second row left figure) is the remote sensing image of August part in Fig. 2, in Fig. 2
4th width figure (the second row right figure) is the remote sensing image of September part, other than seasonal difference, while being had different degrees of
The information such as cloud layer, clouds layer shadow, massif shade, ambient noise.Obtain the vegetation coverage of year one's respective area, any phase
Remotely-sensed data cannot represent the vegetation state in the region completely.
S2 is based on step 1 acquired results, right on the basis of any phase remote sensing image result figure that can cover one's respective area
The remote sensing image of other three phases carries out accurate geometrical registration, makes the remote sensing image data of this four phases in geographical space
Match on coordinate, obtains the remote sensing image result set of graphs of a standard.
S3, for each phase remote sensing image result figure in remote sensing image result set of graphs obtained by step S2 carry out respectively with
Lower processing：
S3-1 calculates separately cloud layer index, vegetation index and shadow index for the phase remote sensing image result figure, obtains
Corresponding index information spatial distribution map；
Since satellite remote sensing sensor is influenced by atmospheric density, cloud layer variation etc., all there is cloud in many images of the same area
Layer occlusion issue.Cloud layer can make the remote sensing terrestrial object information of acquisition decay, or even loss.Cloud cover can reduce in the present embodiment
Or vegetation index information is covered, cause the accuracy of final calculated vegetation coverage, confidence level to be greatly reduced.The present invention passes through
Go deep into cloud layer information satellite remote sensing sensor difference spectral band response characteristic, and with vegetation, soil, water body, bare area,
Cities and towns etc. are in the response similarities and differences of identical band po sition, the spectral band for selecting satellite remote sensing sensor common and shared, structure
One cloud layer index CCI.Cloud layer information can be efficiently extracted based on the cloud layer index, and then eliminates cloud layer information to vegetation
The influence that coverage calculates.The cloud layer index CCI calculation formula are：CCI=(ρR+ρG+ρB)*ρNIR, ρRFor the spectrum of red wave band
Reflectance value, ρGFor the spectral reflectance values of green wave band, ρBFor the spectral reflectance values of blue wave band, ρNIRFor remote sensing image near-infrared
Band spectrum reflectance value.
Currently, normalized differential vegetation index NDVI is widely used in remote sensing monitoring vegetation growth status, while being also this
Embodiment calculates the basis of vegetation coverage.In addition, when to different object spectrum Analysis of response, normalized differential vegetation index
Being introduced into for NDVI can be effectively the atural object area of clouds layer shadow and the low brightness values such as massif shade and Water-Body Information in remote sensing image
It separates.Therefore, normalized differential vegetation index NDVI need to be calculated first, computational methods mainly utilize non-linear normalizing
Method relatively simple ratio vegetation index is handled, calculation formula is：
ρNIRFor remote sensing image near infrared band spectral reflectance values, ρRFor the spectral reflectance values of red wave band.
Cloud layer in addition to terrestrial object information is blocked and weakening effect other than, meanwhile, atural object of cloud layer shade to shadow region itself
Also decrease to some degree acts on vegetation index information.If can be unblanketed by another phase by the terrestrial object information in shadow region
Remote sensing image information is substituted, and the final accuracy for calculating vegetation coverage and confidence level can be preferably improved.In addition to cloud layer
Shade presence is had, because of sun altitude, azimuth and orographic factor, massif shade is also a non-negligible part.
When analyzing the different spectral charactersiticss of targets, the reflectance value of clouds layer shadow and massif shade is very low, if by near infrared band spectrum
Reflectance value (ρNIR) with the spectral reflectance values (ρ of short infrared wave bandSWIR) sum up, shadow information can be widened well
At a distance between other terrestrial object informations, but distance is closer between Water-Body Information and shadow information, be easy to cause misjudgement, the two and not
Shade can preferably be separated with water body；Water body then can be significantly further widened by introducing normalized differential vegetation index (NDVI)
Information between shadow information at a distance from, and then construct a shadow index (YYI), calculation formula is：
Wherein, YYI is shadow index, ρNIRFor remote sensing image near infrared band spectral reflectance values, ρSWIRFor short-wave infrared
The spectral reflectance values of wave band, NDVI are normalized differential vegetation index.
S3-2 calculates background index for the phase remote sensing image result figure, and carries out binary conversion treatment, obtains a shadow
As background information spatial distribution map；
In actual operation, there are many remote sensing images that whole distract can not be completely covered, unlapped region has the back of the body
There is (as shown in Figure 2) in scape information, how to handle the efficiency of the practice that these information will directly affect entire technical method；In addition,
Background information after atmospheric correction by default value be 0, this can cause the vegetation index information value one with other atural objects
Sample, the accuracy that final vegetation coverage calculates will be influenced by dealing with improperly.The present invention is by analyzing different atural objects in each wave
The spectral response characteristics and the order of magnitude of section, construct background index BGI, calculation formula is：
Wherein ρNIRFor remote sensing image near infrared band spectral reflectance values, ρRFor the spectral reflectance values of red wave band, ρGFor
The spectral reflectance values of green wave band.The index effectively can only accurately distinguish background information and terrestrial object information area in image
Come, it is also necessary to further to be handled.The technical program does it further binary conversion treatment, will be carried on the back in background index BGI
Scape information pixel is assigned a value of -99, other pixels are assigned a value of 0, an image background information spatial distribution map BGII is obtained, to make
It participates in calculating in next step for an information factor.
S3-3 carries out cloud algorithm process, the algorithm packet for vegetation index information space distribution map obtained by step S3-1
Include following sub-step：
S3-3-1 first carries out the phase remote sensing image result figure enhancing algorithm process of image before treatment, and
The pixel sample in random acquisition cloud layer area is sample area, the shadow in the remote sensing image result figure after enhancing algorithm process
The enhancing algorithm process of picture refers to 2% Linear Contrast Enhancement Algorithm processing, to increase different type of ground objects color contrasts, it is convenient into
The selection of row next step pixel sample, meanwhile, it selects 2% linear stretch algorithm that can also reduce the influence of environmental background noise, protects
True degree is higher.To avoid mistake from selecting non-overcast area pixel, the random acquisition cloud on the striograph after enhancing algorithm process
The pixel in floor area uses 2 × 2 sampling window random acquisition cloud layer information pixel sample for sample area, to ensure data precision,
Sample pixel quantity is more than 200.
Cloud layer index information spatial distribution map obtained by the sample area and step S3-1 is carried out space overlapping, system by S3-3-2
Meter calculates sample area and corresponds to the cumulative frequency of cloud layer index information numerical value from small to large；
According to the cumulative frequency threshold range is arranged, and make two and cloud layer index information spatial distribution in S3-3-3
The two-value hum pattern that figure matches；The lower limit of the threshold range is cloud layer index information numerical value when cumulative frequency reaches 2%,
The upper limit of the threshold range is infinity.Described two two-value hum patterns to match with cloud layer index information spatial distribution map
It is respectively designated as CCIB1 and CCIB2, CCIB1 is for cloud layer index CCI in step 3, by the numerical value assignment in threshold range
It is 0, other values are assigned a value of 1, calculate and obtain, see in Fig. 3 2.；And CCIB2 is to be directed to cloud layer index CCI, it will be in threshold range
Numerical value is assigned a value of -99, and other values are assigned a value of 0, calculates and obtains, sees in Fig. 3 3..
S3-3-4, by vegetation index information space distribution map and step obtained by two two-value hum patterns of making and step S3-1
Image background information spatial distribution map carries out spatial operation obtained by rapid S3-2, you can the vegetation index after the cloud layer noise that is eliminated
Information result figure；The calculation formula of the spatial operation is：
NDVICCI=NDVI*CCIB1+CCIB2+BGII, wherein NDVICCITo eliminate the vegetation index letter after cloud layer noise
Result figure is ceased, BGII is image background information spatial distribution map, and CCIB1 and CCIB2 are that step S3-3-3 calculates gained two-value letter
Breath figure.More clearly to explain the principle of the spatial operation, contribute to the understanding to numerical procedure of the present invention, if Fig. 3 is the sky
Between calculating process principle schematic diagram.It is 1. wherein NDVI, is 2. CCIB1, is 3. CCIB2, be 4. BGII, be 5. NDVICCI。
S3-4 carries out shadow algorithm and handles for vegetation index information space distribution map obtained by step S3-1, the algorithm
Including following sub-step：
S3-4-1 carries out the phase remote sensing image result figure enhancing algorithm process of image, and by enhancing algorithm
The pixel sample in random acquisition shadow region is sample area in remote sensing image result figure that treated；
Before treatment, the enhancing algorithm process of image is carried out to the phase remote sensing image result figure first, and is being passed through
The pixel sample for enhancing random acquisition shadow region in the remote sensing image result figure after algorithm process is sample area, the increasing of the image
Strong algorithms processing refer to 2% Linear Contrast Enhancement Algorithm processing, to increase different type of ground objects color contrasts, facilitate carry out it is next
Walk the selection of pixel sample, meanwhile, select 2% linear stretch algorithm that can also reduce the influence of environmental background noise, fidelity compared with
It is high.To avoid mistake from selecting nonshaded area pixel, the pixel in random acquisition shadow region on the striograph after enhancing algorithm process
Use 2 × 2 sampling window random acquisition shadow information pixel sample for sample area, to ensure data precision, sample pixel number
Amount is more than 200.
Shadow index information space distribution map obtained by the sample area and step S3-1 is carried out space overlapping, system by S3-4-2
Meter calculates that sample area corresponds to shadow index information value cumulative frequency from small to large and sample area corresponds to shadow index Information Number
Value；
S3-4-3 corresponds to shadow index information value according to the cumulative frequency and sample area, a threshold range is arranged, will tire out
The upper limit of the shadow index information value as threshold value when meter frequency reaches 95%, sample area corresponds to shadow index information value
Lower limit of the minimum value as threshold range, and make two two value informations to match with shadow index information space distribution map
Figure is respectively designated as YYIB1 and YYIB2, and YYIB1 is to be directed to shadow index YYI, and the numerical value in threshold range is assigned a value of 0,
His value is assigned a value of 1, calculates and obtains；YYIB2 is to be directed to shadow index YYI, and the numerical value in threshold range is assigned a value of -99, other
Value is assigned a value of 0, calculates and obtains.
S3-4-4, by vegetation index information data space distribution map obtained by two two-value hum patterns of making and step S3-1
Spatial operation is carried out with image background information spatial distribution map obtained by step S3-2, you can the vegetation after the shade noise that is eliminated
Index information result figure.The calculation formula of the spatial operation is：
NDVIYYI=NDVI*YYIB1+YYIB2+BGII, wherein NDVIYYITo eliminate the vegetation index letter after shadow information
Result figure is ceased, BGII is that step S3-2 calculates gained image background information spatial distribution map, and YYIB1 and YYIB2 are step S3-4-
3 calculate gained two-value hum pattern.The principle of the spatial operation is as in cloud removing algorithm, reference can be made to shown in Fig. 3.
S3-5, based on the image background information spatial distribution map and step S3-3 and step S3-4 made by step S3-2
Made two-value hum pattern makes two new two-value hum patterns, is respectively designated as NDVIB1And NDVIB2, calculation formula
It is as follows respectively：
NDVIB2=(NDVIB1+99)*(NDVIB1-1)
Wherein BGII is that step S3-2 calculates gained background information spatial distribution map, and CCIB1 is that step S3-3 calculates gained
Two-value hum pattern, YYIB1 are that step S3-4 calculates gained two-value hum pattern.
And then new two-value hum pattern is utilized, to the vegetation index information result after elimination cloud layer noise obtained by step S3-3
Figure and step S3-4 gained eliminate the space combination of the vegetation index information result figure progress Mono temporal information after shade noise, raw
At the final vegetation index information result figure of the phase.The space combination method calculation formula is as follows：
Wherein NDVICCIFor the vegetation index information result figure for eliminating after cloud layer noise obtained by step S3-3, NDVIYYIFor step
The vegetation index information result figure after shadow information, NDVI are eliminated obtained by rapid S3-4B1And NDVIB2It is new for two in step S3-5
Two-value hum pattern.
If Fig. 4 is final vegetation index information result figure of four phases through past cloud and after going Shadows Processing, wherein Fig. 4
In the first width figure (the first row left figure) be June final vegetation index information result figure, (the first row is right for the second width figure in Fig. 4
Figure) be July final vegetation index information result figure, third width figure (the second row left figure) is the final vegetation of August part in Fig. 4
Index information result figure, the 4th width figure (the second row right figure) is the final vegetation index information result figure of September part in Fig. 4.
S4, based on the final vegetation index information result figure of 4 phases obtained by step S3, by annual 4 phases in region
Final vegetation index information result figure merged, formed an image data for including multiple wave bands.In the present embodiment
It is to be synthesized the final vegetation index information result figure of four phases obtained by step S3 using the Layer Stacking in ENVI
In the image that one spatially matches.
S5, on the bases step S4, the final vegetation index information result figure of four phases to covering same area,
Using maximum value synthetic method, maximum value synthesis processing is carried out, obtains year regional vegetation index information spatial distribution map, such as figure
5 be maximum value composite result figure.
S6 carries out numerical operation processing to regional vegetation index information spatial distribution map obtained by step S5, eliminates cloud layer, the moon
The influence of shadow and background information to statistic analysis result provides accurately and reliably data source to be for statistical analysis in next step.Tool
Body is that cloud layer information, shadow information and background information are all assigned -99 in above-mentioned calculating process, to avoid these pixels
Information participates in statistics and calculates, and avoids statistical error, needs the area assignment for being -99 by numerical value to be before calculating vegetation coverage
NAN。
S7 is based on step S6 acquired results, and one's respective area vegetative coverage is calculated using the pixel dichotomy based on vegetation index
Degree is to get the accurate vegetation coverage spatial distribution map in the year region, as shown in Figure 6.The pixel dichotomy calculates public
Formula is as follows：
Wherein VFC is the vegetation coverage in pixel, and NDVI is the normalized differential vegetation index on the pixel, NDVIsoilFor
NDVI values without vegetative coverage pixel, NDVIvegFor completely by the pixel NDVI values of vegetative coverage；In the present invention, with 0.5%
The upper lower threshold value of confidence level interception area NDVI respectively represents NDVIvegAnd NDVIsoil, and then zoning vegetation coverage
VFC。
Although reference be made herein to invention has been described for multiple explanatory embodiments of the invention, however, it is to be understood that
Those skilled in the art can be designed that a lot of other modification and implementations, these modifications and implementations will be fallen in this Shen
It please be within disclosed scope and spirit.More specifically, disclose in the application, drawings and claims in the range of, can
With the building block and/or a variety of variations and modifications of layout progress to theme combination layout.In addition to building block and/or layout
Outside the modification and improvement of progress, to those skilled in the art, other purposes also will be apparent.
Claims (22)
1. a kind of year regional vegetation coverage computational methods, which is characterized in that include the following steps：
S1 carries out accurate radiation calibration and atmospheric radiation school to same annual all remote sensing images of vegetation growing season inner region
Just, the remote sensing image result figure that different phases represent atural object real reflectance is obtained；
S2, for remote sensing image result figure obtained by step S1, on the basis of any phase remote sensing image result figure, to other remote sensing
Imaging results figure carries out accurate geometrical registration, forms a remote sensing image result atlas to match on geospatial coordinates
It closes；
S3 carries out following locate respectively for each phase remote sensing image result figure in remote sensing image result set of graphs obtained by step S2
Reason：
S3-1 calculates separately cloud layer index, vegetation index and shadow index for the phase remote sensing image result figure, obtains corresponding
Index information spatial distribution map；
S3-2 calculates background index for the phase remote sensing image result figure, and carries out binary conversion treatment, obtains an image back of the body
Scape information space distribution map；
S3-3 carries out cloud algorithm process for vegetation index information space distribution map obtained by step S3-1, the algorithm include with
Lower sub-step：
S3-3-1 carries out the phase remote sensing image result figure enhancing algorithm process of image, and by enhancing algorithm process
The pixel sample in random acquisition cloud layer area is sample area in remote sensing image result figure afterwards；
Cloud layer index information spatial distribution map obtained by the sample area and step S3-1 is carried out space overlapping, statistics meter by S3-3-2
It calculates sample area and corresponds to the cumulative frequency of cloud layer index information numerical value from small to large；
According to the cumulative frequency threshold range is arranged, and make two and cloud layer index information spatial distribution map phase in S3-3-3
Matched two-value hum pattern；
S3-3-4, by vegetation index information space distribution map and step obtained by two two-value hum patterns of making and step S3-1
Image background information spatial distribution map obtained by S3-2 carries out spatial operation, you can the vegetation index letter after the cloud layer noise that is eliminated
Cease result figure；
S3-4 carries out shadow algorithm and handles, which includes for vegetation index information space distribution map obtained by step S3-1
Following sub-step：
S3-4-1 carries out the phase remote sensing image result figure enhancing algorithm process of image, and by enhancing algorithm process
The pixel sample in random acquisition shadow region is sample area in remote sensing image result figure afterwards；
Shadow index information space distribution map obtained by the sample area and step S3-1 is carried out space overlapping, statistics meter by S3-4-2
Calculate that sample area corresponds to shadow index information value cumulative frequency from small to large and sample area corresponds to shadow index information value；
S3-4-3 corresponds to shadow index information value according to the cumulative frequency and sample area, a threshold range is arranged, and make two
A two-value hum pattern to match with shadow index information space distribution map；
S3-4-4, by vegetation index information data space distribution map and step obtained by two two-value hum patterns of making and step S3-1
Image background information spatial distribution map obtained by rapid S3-2 carries out spatial operation, the vegetation index information after the shade noise that is eliminated
Result figure；
S3-5, based on made by step S3-2 image background information spatial distribution map and step S3-3 and step S3-4 it is made
The two-value hum pattern of work makes two new two-value hum patterns, and then utilizes new two-value hum pattern, to step S3-3 and step
Vegetation index information result figure obtained by S3-4 carries out the space combination of Mono temporal information, generates the final vegetation index letter of the phase
Cease result figure；
S4, based on the final vegetation index information result figure of each phase obtained by step S3, most by annual all phases in region
Whole vegetation index information result figure is merged, and an image data for including multiple wave bands is formed；
S5, to covering the final vegetation index information result figure of same area, is synthesized on the bases step S4 using maximum value
Method carries out maximum value synthesis processing, obtains year regional vegetation index information spatial distribution map；
S6 carries out numerical operation processing to regional vegetation index information spatial distribution map obtained by step S5, eliminate cloud layer, shade and
Influence of the background information to statistic analysis result；
S7 is based on step S6 acquired results, calculates vegetation coverage to get the accurate vegetation coverage space in the year region
Distribution map.
2. a kind of annual regional vegetation coverage computational methods as described in claim 1, which is characterized in that the step S3-1
The cloud layer index CCI of middle structure, calculation formula are：CCI=(ρR+ρG+ρB)*ρNIR, ρRFor the spectral reflectance values of red wave band,
ρGFor the spectral reflectance values of green wave band, ρBFor the spectral reflectance values of blue wave band, ρNIRFor remote sensing image near infrared band spectrum
Reflectance value.
3. a kind of annual regional vegetation coverage computational methods as described in claim 1, which is characterized in that the step S3-1
Middle vegetation index is normalized differential vegetation index NDVI, which is：ρNIRFor remote sensing shadow
As near infrared band spectral reflectance values, ρRFor the spectral reflectance values of red wave band.
4. a kind of annual regional vegetation coverage computational methods as described in claim 1, which is characterized in that the step S3-1
Described in shadow index YYI calculation formula be：Wherein ρNIRFor remote sensing image near infrared band spectrum
Reflectance value, ρSWIRFor the spectral reflectance values of short infrared wave band, NDVI is normalized differential vegetation index.
5. a kind of annual regional vegetation coverage computational methods as described in claim 1, which is characterized in that the step S3-2
The background index BGI of middle structure, calculation formula are：Wherein ρNIRFor remote sensing image near-infrared
Band spectrum reflectance value, ρRFor the spectral reflectance values of red wave band, ρGFor the spectral reflectance values of green wave band.
6. a kind of annual regional vegetation coverage computational methods as described in claim 1, which is characterized in that the step S3-2
Middle binary conversion treatment refers to that background information pixel in background index BGI is assigned a value of -99, other pixels are assigned a value of 0, obtain one
Image background information spatial distribution map BGII.
7. a kind of annual regional vegetation coverage computational methods as described in claim 1, which is characterized in that the step S3-
The enhancing algorithm process of image refers to 2% linear stretch enhancing in 3-1；Random acquisition sample pixel refers to being adopted using 2 × 2
Sample window random acquisition cloud layer information pixel, sample pixel quantity are more than 200.
8. a kind of annual regional vegetation coverage computational methods as described in claim 1, which is characterized in that the step S3-
The setting of threshold range refers to cloud layer index information numerical value when cumulative frequency is reached 2% as the lower limit of threshold value in 3-3, should
Upper threshold limit threshold value is infinity.
9. a kind of annual regional vegetation coverage computational methods as described in claim 1, which is characterized in that the step S3-
Two two-value hum patterns are CCIB1 and CCIB2 in 3-3, and CCIB1 is to be directed to cloud layer index CCI, and the numerical value in threshold range is assigned
Value is 0, and other values are assigned a value of 1, calculates and obtains；CCIB2 is to be directed to cloud layer index CCI, the numerical value in threshold range is assigned a value of-
99, other values are assigned a value of 0, calculate and obtain.
10. a kind of annual regional vegetation coverage computational methods as described in claim 1, which is characterized in that the step S3-
The calculation formula of spatial operation is in 3-4：NDVICCI=NDVI*CCIB1+CCIB2+BGII, wherein NDVICCITo eliminate cloud layer
Vegetation index information result figure after noise, BGII are image background information spatial distribution maps, and CCIB1 and CCIB2 are step S3-
3-3 calculates gained two-value hum pattern.
11. a kind of annual regional vegetation coverage computational methods as described in claim 1, which is characterized in that the step S3-
The enhancing algorithm process of image described in 4-1 refers to 2% linear stretch enhancing；Random acquisition sample pixel refers to using 2 × 2
Sampling window random acquisition image shades information pixel, sample pixel quantity be more than 200.
12. a kind of annual regional vegetation coverage computational methods as described in claim 1, which is characterized in that the step S3-
The setting of threshold range described in 4-3 refers to shadow index information value when cumulative frequency is reached 95% as the upper of threshold value
Limit, sample area corresponds to lower limit of the minimum value as threshold range of shadow index information value.
13. a kind of annual regional vegetation coverage computational methods as described in claim 1, which is characterized in that the step S3-
Two two-value hum patterns described in 4-3 are YYIB1 and YYIB2, and YYIB1 is to be directed to shadow index YYI, by the number in threshold range
Value is assigned a value of 0, and other values are assigned a value of 1, calculates and obtains；YYIB2 is to be directed to shadow index YYI, and the numerical value in threshold range is assigned
Value is -99, and other values are assigned a value of 0, calculates and obtains.
14. a kind of annual regional vegetation coverage computational methods as described in claim 1, which is characterized in that the step S3-
The calculation formula of spatial operation is in 4-4：NDVIYYI=NDVI*YYIB1+YYIB2+BGII, wherein NDVIYYITo eliminate shade
Vegetation index information result figure after information, BGII are that step S3-2 calculates gained image background information spatial distribution map, YYIB1
It is that step S3-4-3 calculates gained two-value hum pattern with YYIB2.
15. a kind of annual regional vegetation coverage computational methods as described in claim 1, which is characterized in that the step S3-
Two new two-value hum patterns are NDVI in 5B1And NDVIB2, calculation formula is distinguished as follows：
NDVIB2=(NDVIB1+99)*(NDVIB1-1)
Wherein BGII is that step S3-2 calculates gained background information spatial distribution map, and CCIB1 is that step S3-3 calculates gained two-value
Hum pattern, YYIB1 are that step S3-4 calculates gained two-value hum pattern.
16. a kind of annual regional vegetation coverage computational methods as described in claim 1, which is characterized in that the step S3-
Space combination method described in 5, calculation formula are as follows：
Wherein NDVICCIFor the vegetation index information result figure for eliminating after cloud layer noise obtained by step S3-3, NDVIYYIFor step
The vegetation index information result figure after shadow information, NDVI are eliminated obtained by S3-4B1And NDVIB2It is new for two in step S3-5
Two-value hum pattern.
17. a kind of annual regional vegetation coverage computational methods as described in claim 1, which is characterized in that institute in step S4
It refers to synthesizing multiple Mono temporal vegetation index information by wave band to obtain an image spatially to match to state fusion.
18. a kind of annual regional vegetation coverage computational methods as described in claim 1, which is characterized in that the step S5
Middle maximum value synthetic method refers to being directed to the same pixel, which has different numerical value in different phases, take in these values
Maximum value, as maximum value synthesize.
19. a kind of annual regional vegetation coverage computational methods as described in claim 1, which is characterized in that institute in step S6
It refers to by numerical value is -99 in regional vegetation index information spatial distribution map obtained by step S5 area assignment to state numerical operation processing
For NAN.
20. a kind of annual regional vegetation coverage computational methods as described in claim 1, which is characterized in that the step S7
The middle vegetation coverage that calculates can estimate the method or model of vegetation coverage using any type based on vegetation index information.
21. a kind of annual regional vegetation coverage computational methods as claimed in claim 20, which is characterized in that described any one
The method or model of kind estimation vegetation coverage include pixel dichotomy, neural network model in Decomposition of Mixed Pixels method.
22. a kind of annual regional vegetation coverage computational methods as claimed in claim 21, which is characterized in that the pixel two
Calculation of group dividing formula is as follows：
Wherein VFC is the vegetation coverage in pixel, and NDVI is the normalized differential vegetation index on the pixel, NDVIsoilFor no vegetation
Cover the NDVI values of pixel, NDVIvegFor completely by the pixel NDVI values of vegetative coverage；With 0.5% confidence level interception area NDVI
Upper lower threshold value respectively represent NDVIvegAnd NDVIsoil, and then zoning vegetation coverage VFC.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711310827.2A CN107909607B (en) | 2017-12-11 | 2017-12-11 | A kind of year regional vegetation coverage computational methods |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711310827.2A CN107909607B (en) | 2017-12-11 | 2017-12-11 | A kind of year regional vegetation coverage computational methods |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107909607A CN107909607A (en) | 2018-04-13 |
| CN107909607B true CN107909607B (en) | 2018-08-03 |
Family
ID=61865133
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711310827.2A Active CN107909607B (en) | 2017-12-11 | 2017-12-11 | A kind of year regional vegetation coverage computational methods |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107909607B (en) |
Families Citing this family (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109164444A (en) * | 2018-09-04 | 2019-01-08 | 中科海慧（北京）科技有限公司 | A kind of natural landscape reconstructing method based on remotely-sensed data |
| CN109238960B (en) * | 2018-09-07 | 2021-02-05 | 兰州大学 | Method for rapidly monitoring grassland actual bearing capacity index based on NDVI |
| CN109344364A (en) * | 2018-10-12 | 2019-02-15 | 江苏师范大学 | A kind of Pixel scrambling parameter calibration method suitable for arid biogeographic zone FVC estimation |
| CN109934109B (en) * | 2019-01-31 | 2022-03-04 | 黄河水利委员会黄河水利科学研究院 | Remote sensing-based method for extracting forest and grass vegetation information in loess plateau water and soil loss area |
| CN109800315B (en) * | 2019-01-31 | 2022-11-29 | 交通运输部天津水运工程科学研究所 | Method for establishing photographic grassland coverage measurement gallery and using method |
| CN110059553B (en) * | 2019-03-13 | 2020-11-10 | 中国科学院遥感与数字地球研究所 | Method for judging and identifying vegetation abnormal features in landslide stage by using optical remote sensing image |
| CN110160992A (en) * | 2019-05-15 | 2019-08-23 | 中国科学院遥感与数字地球研究所 | A kind of Crop classification method based on vegetation coverage time series |
| CN110390267B (en) | 2019-06-25 | 2021-06-01 | 东南大学 | Mountain landscape building extraction method and device based on high-resolution remote sensing image |
| CN110400291A (en) * | 2019-07-02 | 2019-11-01 | 北京师范大学 | The method for extracting A Deli penguin quantity based on RBG color model method |
| CN110427841B (en) * | 2019-07-18 | 2021-09-21 | 中国科学院遥感与数字地球研究所 | Evergreen forest land identification method and system based on remote sensing information |
| CN110378290A (en) * | 2019-07-22 | 2019-10-25 | 中国水利水电科学研究院 | A kind of cloudy optical image data flood Water-Body Information rapid extracting method and system |
| CN110826578B (en) * | 2019-11-14 | 2021-06-18 | 中国地质科学院岩溶地质研究所 | Remote sensing rapid extraction method of karst stony desertification information of global scale |
| CN111126203A (en) * | 2019-12-04 | 2020-05-08 | 山东科技大学 | Dense vegetation identification method based on NDVI percentage matching |
| CN112464167A (en) * | 2020-11-26 | 2021-03-09 | 中国科学院地理科学与资源研究所 | Method for analyzing influence of oil and gas field development on vegetation coverage and landscape pattern |
| CN114689545B (en) * | 2022-03-02 | 2022-11-29 | 珠江水利委员会珠江水利科学研究院 | Vegetation coverage layered estimation method and medium based on DSM (digital surface model) contour slices |
| CN115481368B (en) * | 2022-09-29 | 2023-05-19 | 河北省科学院地理科学研究所 | Vegetation coverage estimation method based on full remote sensing machine learning |
| CN115527123B (en) * | 2022-10-21 | 2023-05-05 | 河北省科学院地理科学研究所 | Land cover remote sensing monitoring method based on multisource feature fusion |
| CN116468321A (en) * | 2023-04-23 | 2023-07-21 | 河北省科学院地理科学研究所 | Topography comprehensive index method for representing vegetation growth distribution rule |
| CN116821589B (en) * | 2023-08-29 | 2023-11-14 | 生态环境部卫星环境应用中心 | Vegetation coverage recovery upper limit calculation method for promoting ecological service function improvement |
| CN116843495B (en) * | 2023-09-01 | 2023-11-10 | 生态环境部卫星环境应用中心 | Method and system for identifying implementation area of vegetation restoration project |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9652691B2 (en) * | 2006-11-07 | 2017-05-16 | The Curators Of The University Of Missouri | Method of predicting crop yield loss due to N-deficiency |
| CN103439757B (en) * | 2013-09-10 | 2015-11-25 | 海全胜 | A kind of cloud detection method of optic utilizing MODIS remote sensing Thermal Infrared Data |
| CN104376204A (en) * | 2014-11-06 | 2015-02-25 | 中国测绘科学研究院 | Method for inverting vegetation coverage by adopting improved pixel dichotomy |
| CN105279738A (en) * | 2015-07-15 | 2016-01-27 | 电子科技大学 | Coping method of shadow problem in vegetation parameter estimation, based on remote sensing images |
| US9898688B2 (en) * | 2016-06-01 | 2018-02-20 | Intel Corporation | Vision enhanced drones for precision farming |
| CN107389036B (en) * | 2017-08-02 | 2019-07-12 | 珠江水利委员会珠江水利科学研究院 | A kind of large spatial scale vegetation coverage calculation method of combination unmanned plane image |
-
2017
- 2017-12-11 CN CN201711310827.2A patent/CN107909607B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN107909607A (en) | 2018-04-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107909607B (en) | A kind of year regional vegetation coverage computational methods | |
| Jia et al. | Fractional vegetation cover estimation algorithm for Chinese GF-1 wide field view data | |
| Yang et al. | Urban land-cover change detection through sub-pixel imperviousness mapping using remotely sensed data | |
| Huang et al. | Jointly assimilating MODIS LAI and ET products into the SWAP model for winter wheat yield estimation | |
| Weng et al. | Generating daily land surface temperature at Landsat resolution by fusing Landsat and MODIS data | |
| Liu et al. | Quantifying spatial–temporal pattern of urban heat island in Beijing: An improved assessment using land surface temperature (LST) time series observations from LANDSAT, MODIS, and Chinese new satellite GaoFen-1 | |
| Essa et al. | Evaluation of the DisTrad thermal sharpening methodology for urban areas | |
| Myeong et al. | A temporal analysis of urban forest carbon storage using remote sensing | |
| Kuc et al. | Sentinel-2 imagery for mapping and monitoring imperviousness in urban areas | |
| Lyu et al. | A new method for grassland degradation monitoring by vegetation species composition using hyperspectral remote sensing | |
| Fu et al. | Estimating landscape net ecosystem exchange at high spatial–temporal resolution based on Landsat data, an improved upscaling model framework, and eddy covariance flux measurements | |
| Deka et al. | Implementation of forest canopy density model to monitor tropical deforestation | |
| Li et al. | Effects of spatial resolution of remotely sensed data on estimating urban impervious surfaces | |
| CN103544477A (en) | Improved linear spectral mixture model based vegetation coverage estimation method | |
| CN115481368B (en) | Vegetation coverage estimation method based on full remote sensing machine learning | |
| Yin et al. | Calculation of land surface emissivity and retrieval of land surface temperature based on a spectral mixing model | |
| 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 | |
| Feng et al. | A hierarchical extraction method of impervious surface based on NDVI thresholding integrated with multispectral and high-resolution remote sensing imageries | |
| Karakacan Kuzucu et al. | Testing the potential of vegetation indices for land use/cover classification using high resolution data | |
| Fitzgerald et al. | Directed sampling using remote sensing with a response surface sampling design for site-specific agriculture | |
| Khoshboresh-Masouleh et al. | A deep learning method for near-real-time cloud and cloud shadow segmentation from gaofen-1 images | |
| Koju et al. | A two-scale approach for estimating forest aboveground biomass with optical remote sensing images in a subtropical forest of Nepal | |
| Stević et al. | Automated identification of land cover type using multispectral satellite images | |
| Bektas Balcik et al. | Determination of magnitude and direction of land use/land cover changes in Terkos Water Basin, Istanbul | |
| Osmond | Hemispherical photography as a tool for urban sustainability evaluation and design |
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 |