CN109427091A - A kind of Biomass Models based on high-spectrum remote-sensing and photogrammetric technology grind construction method - Google Patents

Abstract

The invention discloses a kind of Biomass Models based on high-spectrum remote-sensing and photogrammetric technology to grind construction method.High spectrum resolution remote sensing technique, Unmanned Aerial Vehicle Photogrammetric Technique and terrestrial photogrammetric survey technology is utilized respectively to extract vegetation characteristics parameters, combined ground sample ground measured data, the inverse model between vegetation characteristics parameters and biomass is established respectively, and precision analysis is carried out to the model built is ground, utilize common model-evaluation index coefficient R², the standard error SEE of estimated value, total relative error TRE, average systenatic error MSE, average predictor error MPE the Biomass Models of foundation are needed to evaluate its model, so that it is determined that optimal Biomass retrieval model.The efficiency of inverse process that the present invention grinds established model is preferable, for quickly, it is accurate, comprehensively calculate big region land vehicles biomass and provide new technical method and means.

Description

A kind of Biomass Models based on high-spectrum remote-sensing and photogrammetric technology grind construction method

One, technical field

The present invention relates to a kind of Biomass Models to grind the method built, it is especially a kind of based on high spectrum resolution remote sensing technique, nobodyThe vegetation biomass model of machine photogrammetric technology and terrestrial photogrammetric survey technology grinds construction method.

Two, technical background

Main body of the vegetation as terrestrial ecosystems plays huge in terms of maintaining Carbon balance and keeping species diversityEffect also has great influence to the variation of miniclimate around, and the huge advantage in terms of improving environment has caused to close extensivelyNote.For forest community as maximum land vehicles group, it is raw that the powerful carbon sink of the ecosystem has become international geosphereThe plan of object circle, International Human Dimensions Programme on Global Environmental Change, World Climate Research Program and International Biodiversity plan etc.The research topic of great scientific research plan, and forest carbon storage is calculated by forest biomass conversion at present,Forest biomass is directly or indirectly measured, calculates to obtain forest carbon storage multiplied by percent carbon content in biomass.Therefore, it plantsIt is the basis for assessing Vegetation carbon storage by biomass, forest biomass accounts about 90% or so of land total biomass, and forest is rawObject amount has been classified as one of the important indicator of monitoring forest ecosystem carbon sequestration capacity and carbon remittance amount by world woods connection, how quickly quasi-True acquisition forest biomass and its distribution is of great significance to research organic C storage and carbon cycle.

(1) traditional vegetation biomass estimation is based on fieldwork and investigation vegetation parameter, although this method obtainsThe indexs of correlation such as vegetation biomass and land quality of the survey area obtained accuracy is high, but its process is cumbersome, and the period is long,Effective poor, the man power and material expended in on-site inspection and observation process is larger, it is most important that can cause to vegetation centainlyDestruction, cause this method be appropriate only for small range area vegetation biomass research, to big region or even the whole nation or the whole worldBe restricted when vegetation biomass estimation.

(2) appearance and fast development of contactless remote sensing technology, compensates for the deficiency of traditional estimating and measuring method, especially withoutThe development of man-machine photogrammetric technology and terrestrial photogrammetric survey technology is estimation and the dynamic change prison of big regional vegetation biomassIt surveys and provides a kind of method quickly, convenient and economic, improve the accuracy and rapidity of vegetation biomass estimation, and will notVegetation is damaged.

(3) it is especially the high spectrum resolution remote sensing technique of appearance the 1980s, its main feature is that each pixel not only contains groundObject space information further includes alternative continuous spectral information, according to spectral reflectivity or emissivity curve shape feature,The information such as identification estimation vegetation chlorophyll concentration, vegetation biochemical indicator and vegetation net primary productivity, can using hyperspectral techniqueThe research of qualitative and quantitative is carried out to surface vegetation.

Either traditional biological amount estimation technology or contactless photogrammetric and high spectrum resolution remote sensing technique are in biomassAll there are certain drawbacks in the application in estimation, all do not form the estimation of biomass process and measuring method of system.

Three, summary of the invention

The drawbacks of in order to overcome traditional biological amount field observation mode and existing contactless remote sensing technology biometric measurementExisting limitation is calculated, the object of the present invention is to provide a kind of Biomass Models based on high-spectrum remote-sensing and photogrammetric technologyGrind construction method.

BROAD SUMMARY:

A kind of Biomass Models based on high-spectrum remote-sensing and photogrammetric technology grind construction method, it is characterized in that: benefit respectivelyVegetation characteristics parameters are extracted with high spectrum resolution remote sensing technique, Unmanned Aerial Vehicle Photogrammetric Technique and terrestrial photogrammetric survey technology,Combined ground sample ground measured data, establishes the inverse model between vegetation characteristics parameters and biomass respectively, and to grinding the mould builtType carries out precision analysis, determines optimal Biomass retrieval model.

Wherein, the step of extracting vegetation characteristics parameters using high spectrum resolution remote sensing technique are as follows:

(1) spectroscopic data pre-processes, and selects the average value of the measured value of a certain range before and after a measuring point on the curve of spectrum, makeesFor the value of the point, the sequence { R of N number of measuring point is provided_i, i=1,2,3 ... N }, wherein the value of i point includes each K point in front and backAverage value, utilize mathematical model (1)Wherein, the new value of i point isR′_i, substituted with the average value of the 2k+1 point including this point, and then determine the value of the point on the curve of spectrum；

(2) high-spectral data characteristic parameter extraction utilizes mathematical model (2) FDR_λ(r)=(R_λ(j+1)-R_λ(j))/Δ λ is to bloomModal data is handled, wherein FDR_λ(r)For the reflectivity first derivative at the wavelength i of the midpoint wave band j and j+1, R_λ(j+1)Wave bandValue is the reflectivity of j+1, R_λ(j)Band value is that the reflectivity Δ λ of j is the interval of wavelength j+1 to j, depending on band wavelength, and it is relatedCoefficient is expressed as mathematical model (3)And then obtain the differential curves of multiple orders, can effective green vegetation spectral signature；

(3) mathematical model (4) are utilizedEO-1 hyperion vegetation index model is established,In, R₉₀₀、R₁₀₅₀The wavelength of reflectivity maximum respectively within the scope of near-infrared and short infrared wave band, R₉₅₅、R₁₂₂₀RespectivelyThe wavelength at reflectivity minimum within the scope of near-infrared and short infrared wave band, it is quick between enhanced spectrum feature and vegetation biomassPerception.

Wherein, the detailed process of vegetation characteristics parameters is extracted using Unmanned Aerial Vehicle Photogrammetric Technique are as follows: need surrounding targetThe spiral mode of rising of trees carries out oblique photograph measurement, by oblique photograph measuring technique collect the trees different location,The image data of different angle rebuilds the three-dimensional point cloud model of ground vegetation, wherein between the adjacent image of sustained height acquisitionReach 90% or more Duplication, reaches 60% or more Duplication between the image of different height acquisition, phase when oblique photographMachine best angle is 45 degree；According to feature in three dimensional point cloud successively selected characteristic point, using Related Mathematical Models respectively intoRow tree height and the measuring and calculating of trunk any place diameter, crown mapping area and the measuring and calculating of hat width, the measuring and calculating of Crown surface area, Tree Crown VolumeMeasuring and calculating.

Wherein, the step of extracting vegetation characteristics parameters using terrestrial photogrammetric survey technology are as follows: with common optics or numberCamera is tool, establishes digital close view photogrammetric system by computer, and system photographs to permanent sample plot by general camera,Within foundation, elements of exterior orientation (x₀, y₀, f, x_s, y_s, z_s, X, K, U) and it is the photogrammetric collinearity equation of unknown number, become by DLTIt changes, undetermined parameter is acquired, and then image space and object space are connected, by computer system, so that it may determine image spaceThe corresponding object coordinates of any point coordinate (x, y) (X, Y, Z) calculate tree height, the diameter of a cross-section of a tree trunk 1.3 meters above the ground, tree crown to restore object space solidVolume surveys tree index.

Wherein, Biomass Models, which are ground, builds detailed process are as follows:

(1) common Biomass Models W=aD is utilized^b, W=a (D²H)^b, W=aD^bH^c、Wherein, W indicates that biomass, D indicate the diameter of a cross-section of a tree trunk 1.3 meters above the ground, and H indicates that tree is high, C_wIndicate hat width, a and b are estimation parameter, by EO-1 hyperion, nobodyThe individual tree information that machine, terrestrial photogrammetric survey technology are extracted establishes Biomass retrieval model in conjunction with the measured data of ground；

(2) common model-evaluation index coefficient R is utilized², the standard error SEE of estimated value, total relative errorTRE, average systenatic error MSE, average predictor error MPE need to evaluate its model to the Biomass Models of foundation,To select optimal models.

This invention has the advantage that compared with the conventional method

(1) change traditional biological amount and calculate mode on the spot, reduce field process amount, improve measurement efficiency；

(2) in terms of measurement range, big region is promoted to by original small range measuring and calculating and quickly, is efficiently calculated；

(3) evade the limitation of the prior art, mutually use for reference and utilize, realize accurate, the efficient measuring and calculating of biomass.

Four, Detailed description of the invention

The present invention is further described with example with reference to the accompanying drawing.

Fig. 1 is that this technology invention biomass calculates framework map；

Fig. 2 is unmanned plane oblique photograph schematic diagram；

Fig. 3 a is that tree crown microtomy is sliced schematic diagram；

Fig. 3 b is that tree crown microtomy measuring and calculating point chooses schematic diagram；

Fig. 4 a is that crown mapping method extracts schematic diagram；

Fig. 4 b is that crown mapping method extracts schematic diagram；

Fig. 5 is terrestrial photogrammetric survey tree height measurement schematic diagram；

Fig. 6 is terrestrial photogrammetric survey tree breast diameter survey schematic diagram；

Fig. 7 is that Tree Crown Volume split plot design seeks calculation schematic diagram.

Five, specific embodiment:

The present invention is further illustrated With reference to embodiment, specifically:

Vegetation characteristics parameters are extracted using high spectrum resolution remote sensing technique, grind the step of building Biomass Models are as follows:

(1) spectroscopic data pre-processes, and selects the average value of the measured value of a certain range before and after a measuring point on the curve of spectrum, makeesFor the value of the point, the sequence { R of N number of measuring point is provided_i, i=1,2,3 ... N }, wherein the value of i point includes each K point in front and backAverage value, utilize mathematical model (1)Wherein, the new value of i point isR′_i, substituted with the average value of the 2k+1 point including this point, and then determine the value of the point on the curve of spectrum；

(2) high-spectral data characteristic parameter extraction utilizes mathematical model (2) FDR_λ(r)=(R_λ(j+1)-R_λ(j))/Δ λ is to bloomModal data is handled, wherein FDR_λ(r)For the reflectivity first derivative at the wavelength i of the midpoint wave band j and j+1, R_λ(j+1)Wave bandValue is the reflectivity of j+1, R_λ(j)Band value is that the reflectivity Δ λ of j is the interval of wavelength j+1 to j, depending on band wavelength, and it is relatedCoefficient is expressed as mathematical model (3)And then obtain the differential curves of multiple orders, can effective green vegetation spectral signature；

(3) mathematical model (4) are utilizedEO-1 hyperion vegetation index model is established,In, R₉₀₀、R₁₀₅₀The wavelength of reflectivity maximum respectively within the scope of near-infrared and short infrared wave band, R₉₅₅、R₁₂₂₀RespectivelyThe wavelength at reflectivity minimum within the scope of near-infrared and short infrared wave band, it is quick between enhanced spectrum feature and vegetation biomassPerception；

(4) common Biomass Models W=aD is utilized^b, W=a (D²H)^b, W=aD^bH^c、Wherein, W indicates that biomass, D indicate the diameter of a cross-section of a tree trunk 1.3 meters above the ground, and H indicates that tree is high, C_wIndicate hat width, a and b are estimation parameter, by EO-1 hyperion, nobodyThe individual tree information that machine, terrestrial photogrammetric survey technology are extracted establishes Biomass retrieval model in conjunction with the measured data of ground；

(5) common model-evaluation index coefficient R is utilized², the standard error SEE of estimated value, total relative errorTRE, average systenatic error MSE, average predictor error MPE need to evaluate its model to the Biomass Models of foundation,To select optimal models.

Vegetation characteristics parameters are extracted using Unmanned Aerial Vehicle Photogrammetric Technique, grind the step of building Biomass Models are as follows:

(1) as shown in Fig. 2, surrounding target trees spiral, the mode of rising carries out oblique photograph measurement, passes through oblique photographMeasuring technique collects the image data of the trees different location, different angle, rebuilds the three-dimensional point cloud model of ground vegetation,In, 90% or more Duplication is reached between the adjacent image of sustained height acquisition, is reached between the image of different height acquisition60% or more Duplication, camera best angle when oblique photograph are 45 degree；

(2) tree height and the detailed process of trunk any place diameter measuring and calculating are as follows:

1. i indicates point number, value 1,2 ..., n, if certain point i on trunk on the three dimensional point cloud of target treesCoordinate be (X_i, Y_i, Z_i), the coordinate of another point i+1 is (X_i+1, Y_i+1, Z_i+1), it utilizes mathematical model (5)Acquire each Duan Shugao, wherein H indicates each section of tree high level, and i and i+1 is respectively about two points on the same vertical line of trunk at this time；

2. utilizing mathematical model (6)Acquire any place tree-walk value, wherein D is indicatedTree-walk value, i and i+1 is respectively two left-right dots in trunk same level at this time.

(3) detailed process of the measuring and calculating of crown mapping area and hat width are as follows:

1. the three dimensional point cloud of target trees is carried out hierarchy slicing, as shown in Fig. 3 a, 3b, if i-th on a certain sliceThe coordinate of a point is (X_i, Y_i), the coordinate of i+1 point is (X_i+1, Y_i+1), then utilize mathematical model (7)Calculate slice area, wherein i indicate point number, value 1,2 ...,n；

2., when slice area reaches maximum, maximum at this time is sliced area S as shown in Fig. 4 a, 4b_maxMean that tree crownProjected area；

3. utilizing mathematical model (8)Acquire the hat width of tree crown, wherein K indicates hatWidth, (X_a, Y_a)、(X_b, Y_b) respectively indicate lie farthest away on crown mapping section two o'clock a and b coordinate.

(4) detailed process of the measuring and calculating of Crown surface area are as follows:

1. utilizing mathematical model (9)It is respectively sliced after calculating tree crown layeringPerimeter, wherein (X_i, Y_i) it is to be sliced upper i-th point of coordinate, (X_i+1, Y_i+1) be i+1 point coordinate, i expression point number, takeValue is 1,2 ..., n；

It is made of 2. regarding as the tree crown after layering several intermediate stage bodies and top and the cone of bottom two, stage bodySide expansion be approximately it is trapezoidal, the expansion of the side of centrum is approximately triangle, is utilized respectively mathematical model (10)(11)Calculate the lateral area of approximate trapezoid and triangle, wherein j expression is cutPiece number, value are 1 ..., m, and when indicating stage body slice, value is 2 ..., m-1, and when indicating cone, value is 1 or m, L_j、L_j+1Respectively indicate the slice perimeter of jth layer and+1 layer of jth, h_jIndicate that the side of folded rule body between adjacent two layers are sliced is high；

3. the lateral area of each rule body acquired is added up, utilize mathematical model (12)Acquire tree crown surfaceProduct, wherein Q is Crown surface area.

(5) detailed process of the measuring and calculating of Tree Crown Volume are as follows:

1. tree crown is layered, the tree crown after layering is regarded as by several intermediate stage bodies and top and the cone of bottom twoIt constitutes, utilizes mathematical model (13)(14)Calculate approximate platformThe volume of body and cone, wherein j indicates that slice number, value are 1 ..., m, and value is 2 ..., m- when indicating stage body slice1, when indicating cone, value is 1 or m, S_j、S_j+1Respectively indicate the slice area of jth layer and+1 layer of jth, H_jIndicate adjacent two layersSegmentation height between slice；

2. the volume of each rule body acquired is added up, utilize mathematical model (15)Crown surface area is acquired,Wherein, V is Crown surface area；

(6) common Biomass Models W=aD is utilized^b, W=a (D²H)^b, W=aD^bH^c、Wherein, W indicates that biomass, D indicate the diameter of a cross-section of a tree trunk 1.3 meters above the ground, and H indicates that tree is high, C_wIndicate hat width, a and b are estimation parameter, by EO-1 hyperion, nobodyThe individual tree information that machine, terrestrial photogrammetric survey technology are extracted establishes Biomass retrieval model in conjunction with the measured data of ground；

(7) common model-evaluation index coefficient R is utilized², the standard error SEE of estimated value, total relative errorTRE, average systenatic error MSE, average predictor error MPE need to evaluate its model to the Biomass Models of foundation,To select optimal models.

Vegetation characteristics parameters are extracted using terrestrial photogrammetric survey technology, grind the step of building Biomass Models are as follows:

(1) using common optics or digital camera as tool, digital close view photogrammetric system is established by computer, isSystem photographs to permanent sample plot by general camera, within foundation, elements of exterior orientation (x₀, y₀, f, x_s, y_s, z_s, X, K, U) and it is unknownThe photogrammetric collinearity equation of number, is converted by DLT, acquires undetermined parameter, and then image space and object space are connected,By computer system, so that it may the corresponding object coordinates of image space any point coordinate (x, y) (X, Y, Z) are determined, to restore object spaceSpace multistory；

(2) specific steps of the high measuring and calculating of tree are as follows:

1. the flat position apart from the high left and right of 1 times of object to be measured tree tree is selected to guarantee goal tree and instrument as observation pointBetween intervisibility it is unobstructed；

2. frame level device, and measure the high h of instrument at this time；

3. as shown in figure 5, alignment target tree, carries out horizontal honest shooting.

(3) specific steps of diameter of a cross-section of a tree trunk 1.3 meters above the ground measuring and calculating are as follows:

1. selecting suitable observation point, frame level device measures height of instrument at this time；

2. as shown in fig. 6, screen center alignment target tree is carried out horizontal honest shooting, dragging photo horizontal line V and groundUpper thread V_gIt is overlapped, the anti-diameter of a cross-section of a tree trunk 1.3 meters above the ground horizontal line released at 1.3m；

3. it is left along coordinate (U to choose the 1.3m diameter of a cross-section of a tree trunk 1.3 meters above the ground manually_l, V_1.3) and it is right along coordinate (U_r, V_1.3)；

4. utilizing mathematical model (16)Calculate the diameter of a cross-section of a tree trunk 1.3 meters above the ground d of goal tree.

(4) specific steps of Tree Crown Volume measuring and calculating are as follows:

1. setting height-age curve domain as D, tree crown is split with bis- straight line of X, Y is parallel to, tree crown is divided into n notEquitant daughter, as shown in Figure 7；

2. calculating the Tree Crown Volume that each daughter is controlled, and add up as n → ∞, mathematical model (17) V=∫ ∫ is utilizedH_id_xd_y, acquire entire Tree Crown Volume, wherein d_xd_yFor the floor space of daughter,

(5) common Biomass Models W=aD is utilized^b, W=a (D²H)^b, W=aD^bH^c、Wherein, W indicates that biomass, D indicate the diameter of a cross-section of a tree trunk 1.3 meters above the ground, and H indicates that tree is high, C_wIndicate hat width, a and b are estimation parameter, by EO-1 hyperion, nobodyThe individual tree information that machine, terrestrial photogrammetric survey technology are extracted establishes Biomass retrieval model in conjunction with the measured data of ground；

(6) common model-evaluation index coefficient R is utilized², the standard error SEE of estimated value, total relative errorTRE, average systenatic error MSE, average predictor error MPE need to evaluate its model to the Biomass Models of foundation,To select optimal models.

Claims (12)

1. a kind of Biomass Models based on high-spectrum remote-sensing and photogrammetric technology grind construction method, it is characterized in that: being utilized respectivelyHigh spectrum resolution remote sensing technique, Unmanned Aerial Vehicle Photogrammetric Technique and terrestrial photogrammetric survey technology extract vegetation characteristics parameters, knotGround sample ground measured data is closed, establishes the inverse model between vegetation characteristics parameters and biomass respectively, and to grinding the model builtPrecision analysis is carried out, determines optimal Biomass retrieval model.

2. the Biomass Models according to claim 1 based on high-spectrum remote-sensing and photogrammetric technology grind construction method,The step of being characterized in that, extracting vegetation characteristics parameters using high spectrum resolution remote sensing technique are as follows:

(1) spectroscopic data pre-processes, and the average value of the measured value of a certain range before and after a measuring point on the curve of spectrum is selected, as thisThe value of point, provides the sequence { R of N number of measuring point_i, i=1,2,3 ... N }, wherein the value of i point includes the flat of each K point in front and backMean value utilizes mathematical model (1)Wherein, the new value of i point is R '_i,It is substituted with the average value of the 2k+1 point including this point, and then determines the value of the point on the curve of spectrum；

(2) high-spectral data characteristic parameter extraction utilizes mathematical model (2) FDR_λ(r)=(R_λ(j+1)-R_λ(j))/Δ λ is to EO-1 hyperion numberAccording to being handled, wherein FDR_λ(r)For the reflectivity first derivative at the wavelength i of the midpoint wave band j and j+1, R_λ(j+1)Band valueFor the reflectivity of j+1, R_λ(j)Band value is that the reflectivity Δ λ of j is the interval of wavelength j+1 to j, depending on band wavelength, and it is relatedCoefficient is expressed as mathematical model (3)And then obtain the differential curves of multiple orders, can effective green vegetation spectral signature；

(3) mathematical model (4) are utilizedEstablish EO-1 hyperion vegetation index model, whereinR₉₀₀、R₁₀₅₀The wavelength of reflectivity maximum respectively within the scope of near-infrared and short infrared wave band, R₉₅₅、R₁₂₂₀It is respectively closeThe wavelength at reflectivity minimum within the scope of infrared and short infrared wave band, it is sensitive between enhanced spectrum feature and vegetation biomassProperty.

3. the Biomass Models according to claim 1 based on high-spectrum remote-sensing and photogrammetric technology grind construction method,It is characterized in that, extracts vegetation characteristics parameters using Unmanned Aerial Vehicle Photogrammetric Technique, surrounding target trees is needed to spiral the side of risingFormula carries out oblique photograph measurement, and the image number of the trees different location, different angle is collected by oblique photograph measuring techniqueAccording to rebuilding the three-dimensional point cloud model of ground vegetation, wherein reach 90% or more weight between the adjacent image of sustained height acquisitionFolded rate, reaches 60% or more Duplication between the image of different height acquisition, and camera best angle when oblique photograph is 45 degree；According to feature in three dimensional point cloud successively selected characteristic point, carry out that tree is high and trunk any place respectively using Related Mathematical ModelsDiameter measuring and calculating, crown mapping area and the measuring and calculating of hat width, the measuring and calculating of Crown surface area, the measuring and calculating of Tree Crown Volume.

4. the method that Unmanned Aerial Vehicle Photogrammetric Technique according to claim 3 extracts vegetation characteristics parameters, which is characterized in thatTree height and the detailed process of trunk any place diameter measuring and calculating are as follows:

(1) on the three dimensional point cloud of target trees, i indicates point number, value 1,2 ..., n, if on trunk certain point i seatIt is designated as (X_i, Y_i, Z_i), the coordinate of another point i+1 is (X_i+1, Y_i+1, Z_i+1), it utilizes mathematical model (5)It asksObtain each Duan Shugao, wherein H indicates each section of tree high level, and i and i+1 is respectively about two points on the same vertical line of trunk at this time；

(2) mathematical model (6) are utilizedAcquire any place tree-walk value, wherein D indicates tree-walkValue, i and i+1 is respectively two left-right dots in trunk same level at this time.

5. the method that Unmanned Aerial Vehicle Photogrammetric Technique according to claim 3 extracts vegetation characteristics parameters, which is characterized in thatThe detailed process of the measuring and calculating of crown mapping area and hat width are as follows:

(1) three dimensional point cloud of target trees is subjected to hierarchy slicing, if a certain upper i-th point of the coordinate of slice is (X_i, Y_i),The coordinate of i+1 point is (X_i+1, Y_i+1), then utilize mathematical model (7)MeterCalculate the area of slice, wherein i indicates point number, value 1,2 ..., n；

(2) when slice area reaches maximum, maximum at this time is sliced area S_maxMean that crown mapping area；

(3) mathematical model (8) are utilizedAcquiring the hat width of tree crown, wherein K indicates hat width,(X_a, Y_a)、(X_b, Y_b) respectively indicate lie farthest away on crown mapping section two o'clock a and b coordinate.

6. the method that Unmanned Aerial Vehicle Photogrammetric Technique according to claim 3 extracts vegetation characteristics parameters, which is characterized in thatThe detailed process of the measuring and calculating of Crown surface area are as follows:

(1) mathematical model (9) are utilizedCalculate the week being respectively sliced after tree crown layeringIt is long, wherein (X_i, Y_i) it is to be sliced upper i-th point of coordinate, (X_i+1, Y_i+1) be i+1 point coordinate, i expression point number, valueIt is 1,2 ..., n；

(2) tree crown after layering is regarded as and is made of several intermediate stage bodies and top and the cone of bottom two, the side exhibition of stage bodyIt is approximately trapezoidal for opening, and the side expansion of centrum is approximately triangle, is utilized respectively mathematical model (10)(11)Calculating the lateral area of approximate trapezoid and triangle, wherein j indicates that slice number, value are 1 ..., m,When indicating stage body slice, value is 2 ..., m-1, and when indicating cone, value is 1 or m, L_j、L_j+1Respectively indicate jth layer and+ 1 layer of jth of slice perimeter, h_jIndicate that the side of folded rule body between adjacent two layers are sliced is high；

(3) lateral area of each rule body acquired is added up, is utilized mathematical model (12)Crown surface area is acquired,Wherein, Q is Crown surface area.

7. the method that Unmanned Aerial Vehicle Photogrammetric Technique according to claim 3 extracts vegetation characteristics parameters, which is characterized in thatThe detailed process of the measuring and calculating of Tree Crown Volume are as follows:

(1) tree crown is layered, the tree crown after layering is regarded as by several intermediate stage bodies and top and the cone structure of bottom twoAt utilizing mathematical model (13)(14)Calculate approximate stage bodyWith the volume of cone, wherein j indicates that slice number, value are 1 ..., m, and when indicating stage body slice, value is 2 ..., m-1,When indicating cone, value is 1 or m, S_j、S_j+1Respectively indicate the slice area of jth layer and+1 layer of jth, H_jIndicate that adjacent two layers are cutSegmentation height between piece；

(2) volume of each rule body acquired is added up, is utilized mathematical model (15)Crown surface area is acquired,In, V is Crown surface area.

8. the Biomass Models according to claim 1 based on high-spectrum remote-sensing and photogrammetric technology grind construction method,The step of being characterized in that, extracting vegetation characteristics parameters using terrestrial photogrammetric survey technology are as follows: with common optics or digital cameraFor tool, digital close view photogrammetric system is established by computer, system is photographed to permanent sample plot by general camera, establishedWithin, elements of exterior orientation (x₀, y₀, f, x_s, y_s, z_s, X, K, U) and it is the photogrammetric collinearity equation of unknown number, it is converted, is asked by DLTUndetermined parameter is obtained, and then image space and object space are connected, by computer system, so that it may determine image space any pointThe corresponding object coordinates of coordinate (x, y) (X, Y, Z) calculate tree height, the diameter of a cross-section of a tree trunk 1.3 meters above the ground, Tree Crown Volume and survey to restore object space solidSet index.

9. the method that terrestrial photogrammetric survey technology according to claim 8 extracts vegetation characteristics parameters, which is characterized in that treeThe specific steps of height measuring and calculating are as follows:

(1) flat position apart from the high left and right of 1 times of object to be measured tree tree is selected to guarantee between goal tree and instrument as observation pointIntervisibility is unobstructed；

(2) frame level device, and measure the high h of instrument at this time；

(3) alignment target tree carries out horizontal honest shooting.

10. the method that terrestrial photogrammetric survey technology according to claim 8 extracts vegetation characteristics parameters, which is characterized in thatThe specific steps of diameter of a cross-section of a tree trunk 1.3 meters above the ground measuring and calculating are as follows:

(1) suitable observation point is selected, frame level device measures height of instrument at this time；

(2) by screen center alignment target tree, horizontal honest shooting, dragging photo horizontal line V and ground line V are carried out_gIt is overlapped, insteadRelease the diameter of a cross-section of a tree trunk 1.3 meters above the ground horizontal line at 1.3m；

(3) it is left along coordinate (U that the 1.3m diameter of a cross-section of a tree trunk 1.3 meters above the ground is chosen manually_l, V_1.3) and it is right along coordinate (U_r, V_1.3)；

(4) mathematical model (16) are utilizedCalculate the diameter of a cross-section of a tree trunk 1.3 meters above the ground d of goal tree.

11. the method that terrestrial photogrammetric survey technology according to claim 8 extracts vegetation characteristics parameters, which is characterized in thatThe specific steps of Tree Crown Volume measuring and calculating are as follows:

(1) height-age curve domain is set as D, tree crown is split with bis- straight line of X, Y is parallel to, and tree crown is divided into n not phasesThe daughter of overlapping；

(2) as n → ∞, the Tree Crown Volume that each daughter is controlled is calculated, and add up, utilizes mathematical model (17) V=∫ ∫H_id_xd_y, acquire entire Tree Crown Volume, wherein d_xd_yFor the floor space of daughter,

12. the Biomass Models according to claim 1 based on high-spectrum remote-sensing and photogrammetric technology grind construction method,It is characterized in that, Biomass Models, which are ground, builds detailed process are as follows:

(1) common Biomass Models W=aD is utilized^b, W=a (D²H)^b, W=aD^bH^c、Wherein,W indicates that biomass, D indicate the diameter of a cross-section of a tree trunk 1.3 meters above the ground, and H indicates that tree is high, C_wIndicate hat width, a and b are estimation parameter, by EO-1 hyperion, unmanned plane,The individual tree information that face photogrammetric technology is extracted establishes Biomass retrieval model in conjunction with the measured data of ground；

(2) common model-evaluation index coefficient R is utilized², it is the standard error SEE of estimated value, total relative error TRE, averageSystematic error MSE, average predictor error MPE need to evaluate its model, to select to the Biomass Models of foundationOptimal models.