CN106124049A - A kind of implementation method of Vegetation canopy multi-optical spectrum imaging system - Google Patents

Abstract

The present invention relates to the implementation method of a kind of Vegetation canopy multi-optical spectrum imaging system, first transformation slr camera makes it possess near infrared band imaging capability, camera is placed in The Cloud Terrace and spider, rotary platform is to setting zenith angle and azimuth direction, fish eye lens or focal length, tight shot are installed, and filter is installed on camera lens, make camera at the single wave band of near-infrared or visible ray and two wave bands of near-infrared or multi-spectral imaging, complete multiple zenith angle and azimuth direction canopy multi-spectra imaging according to setting scheme.Afterwards the Vegetation canopy multispectral image of collection is classified, and based on classification image, calculate the wooden gross area of Vegetation canopy than the distribution of parameter, aggregate index and photosynthetically active radiation.The present invention, without the sampling of Vegetation canopy destructiveness, manpower and materials are put into the problems such as little, can to solve the existence of traditional optical measuring method error is big, efficiency is low, can be applicable to the wooden gross area of Vegetation canopy than parameter, aggregate index and photosynthetically active radiation distribution measuring.

Description

A kind of implementation method of Vegetation canopy multi-optical spectrum imaging system

Technical field

The present invention relates to Vegetation canopy multi-optical spectrum imaging technology field, particularly relate to a kind of Vegetation canopy multispectral imaging The implementation method of system.

Background technology

Leaf area index (leaf area index, LAI) is dimensionless, and it may be defined as on unit surface area all The half that green plants organ surface is long-pending.As one of core parameter characterizing Vegetation canopy structure, current LAI has been widely used In models such as vegetation growth and productivity model, crop growth model, Net Primary Productivity Model, Atmospheric models, hydrological models And the field such as forestry, botany, ecology, agronomy.

Photosynthetically active radiation is as the main energy sources of synthetic carbohydrate in photosynthesis of plant, and it is in plant canopy Distribution can affect the change of biological yield of plant, may result in again the change of phytomorph structure.Therefore Vegetation canopy light Close Net long wave radiation distribution and can be widely applied to the fields such as forestry, botany, agronomy.

Vegetation canopy ground LAI measuring method is broadly divided into direct measuring method and indirect measurement method.The side of measurement indirectly Method includes relative growth algoscopy, Leaf-area index and optical measuring method.Optical measurement is mainly used in the middle of current practice Method measures Vegetation canopy LAI, and the instrument that this measuring method is used mainly includes LAI-2000, HemiView, TRAC, DHP (Digital Hemispheral Photography, hemisphere method for imaging) and SunScan etc..Numerous studies show, and directly Measuring method is compared, and LAI measuring method would generally underestimate LAI about 20%～50%.Ground LAI measuring method is measured Error is mainly derived from three source of errors, and the most wooden gross area is than parameter measurement, aggregate index measurement, orographic effect, therefore Carry out the wooden gross area of Vegetation canopy to measure for improving the indirect certainty of measurement of ground LAI and reliability than parameter and aggregate index The most important.

Tradition LAI measuring method theoretical model is clearance rate model, and this model hypothesis canopy solvent is mixed Turbid medium, i.e. spatial distribution are random distribution.The actually solvent spatial distribution of overwhelming majority Vegetation canopy is not the most random, There is building-up effect in i.e. canopy solvent.Internal two yardsticks of Vegetation canopy are (in canopy solvent and canopy solvent Portion) all there is building-up effect, thus canopy solvent building-up effect and canopy solvent accumulated inside effect can be divided into, because of And aggregate index can be divided into canopy solvent aggregate index (Ω e) and wood components aggregate index (Ω w).The most conventional Aggregate index computational algorithm has the distribution of gap length Distribution Algorithm, finite length averaging method, segregation coefficient method, gap length and has Limit for length's degree average combination method scheduling algorithm.Owing to the most traditional measuring method does not possess, Vegetation canopy is become at multiple wave bands Picture, therefore it cannot realize the measurement of Vegetation canopy wood components aggregate index, and wood components aggregate index is measured for vegetation The wooden area index of canopy is accurately measured and most important.

Traditional Forest Canopy LAI measuring method all cannot effectively distinguish Vegetation canopy wood components (trunk, branch, really The non-photosynthetic effect components such as reality) and canopy solvent (leaves), thus cause its measurement result to be gross area index (Plant area index).Owing to only needing Vegetation canopy LAI parameter during actual application, rather than PAI parameter, thus cause passing System measuring method is often difficult to directly meet practical application request.By combining Vegetation canopy multispectral image, based on necessarily Measurement scheme and computation model can calculate the wooden gross area of Vegetation canopy than parameter, thus realize traditional optical measuring method Measurement result (PAI) be converted to the LAI needed for real world applications.

Traditional Vegetation canopy photosynthetically active radiation is measured and is often calculated based on flake image (visible light wave range), due to flake Image (visible light wave range) is collected easily to be affected by weather condition, and flake image itself is easier to over-exposed simultaneously, thus causes hat Layer light and Net long wave radiation computational accuracy reduce.

Summary of the invention

In view of this, it is an object of the invention to provide the implementation method of a kind of Vegetation canopy multi-optical spectrum imaging system, the party Method can to Vegetation canopy High Accuracy Observation and measurement, system cost is low, field work efficiency high.

The present invention uses below scheme to realize: the implementation method of a kind of Vegetation canopy multi-optical spectrum imaging system, including following Step:

Step S1: transform a slr camera so that it is can be near infrared band imaging；

Step S2: be placed on The Cloud Terrace and spider by described slr camera, rotary platform is to setting zenith angle and orientation Angular direction, installs fish eye lens, telephoto lens or tight shot, and installs filter on camera lens；

Step S3: make described slr camera at the single wave band of near-infrared or visible ray and two wave bands of near-infrared or multiband Imaging, completes episphere direction Vegetation canopy multispectral imaging；

Step S4: the Vegetation canopy multispectral image of collection is classified；

Step S5: calculate the wooden gross area of Vegetation canopy based on image of classifying than parameter, aggregate index and the sun Photosynthetically active radiation is distributed.

Further, in described step S1, described slr camera include Nikon D300, Nikon D80, Canon 5D, Canon 70D, Canon 700D, Nikon D7100, micro-one camera, and other type of camera imaging system.

Further, in described step S1, it is by internal for slr camera CCD original paper that the machine that described list is anti-phase carries out transformation Anterior IR-cut filter replaces with near-infrared filter, and described IR-cut filter is low passage filter.

Further, in described step S2, described The Cloud Terrace realizes 0 ° of-360 ° of direction at azimuth direction and rotates, at zenith Angular direction realizes rotating in 0 ° of-180 ° of direction.

Further, in described step S3, described visible ray and two wave band imagings of near-infrared are by filtering before replacing camera lens The method of sheet realizes drift angle and azimuth direction Vegetation canopy multispectral imaging on the same day.

Further, in described step S3, when the camera lens that slr camera carries is fish eye lens, described slr camera mirror Head direction should be towards zenith direction, and described slr camera single imaging can realize episphere direction Vegetation canopy multispectral imaging； When the camera lens that slr camera camera carries is focal length or tight shot, by half direction of bowl in zenith angle and azimuth direction difference Being divided into several zenith angles and azimuth angle interval, described slr camera is all planted in each zenith angle and azimuth direction By canopy multi-spectra imaging.

Further, in described step S4, the classification of described Vegetation canopy multispectral image uses supervised classification or point supervision Sorting technique is to Vegetation canopy visible ray and near infrared band image classification, then Vegetation canopy image is divided into wood components, leaves And sky three major types.

Further, in described step S5, the wooden gross area of described Vegetation canopy than parameter technology with classification image on Based on the clearance rate of wood components, leaves and sky component, calculate sampled point PAI and WAI according to Beer law computational methods, And then derive the wooden gross area of sampled point Vegetation canopy than parameter, concrete calculation procedure is as follows:

Step S511: zenith angle direction is in turn divided into 9 zenith angle intervals, i.e. (0 °-5 °), (5 °-15 °), (15 °- 25 °), (25 °-35 °), (35 °-45 °), (45 °-55 °), (55 °-65 °), (> 65 °-75 °), (75 °-85 °), statistics is each respectively Sky, leaves, wood components and the number of pixels of soil constitution on individual classification image, each azimuth side interval to each zenith angle On classification image, sky, leaves, the number of pixels of wood components and soil constitution add up, obtain 9 zenith angle direction skies, The sum of all pixels of leaves, wood components and soil constitution；

Step S512: each zenith angle interval sky, leaves and the pixel summation of soil constitution and the ratio of image pixel sum Value is the clearance rate p (θ of wood components_wi), and the ratio of the pixel summation of sky and image pixel sum to be canopy basic Clearance rate p (the θ of component_pi), effective gross area indices P AI that each zenith angle is interval_eiWith wooden area index WAI_eiComputational methods As follows:

PAI_ei=-ln [p (θ_pi)]cos(θ_i), W_i=sin (θ_i)d(θ_i)

WAI_ei=-ln [p (θ_wi)]cos(θ_i), W_i=sin (θ_i)d(θ_i)

In formula, i is zenith angle Interval Coding, θ_iFor zenith angle；

Sampled point effective gross area indices P AI_eWith wooden area index WAI_eComputational methods are as follows:

${\mathrm{PAI}}_e=2\underset{i=1}{\overset{9}{\Σ}}{\mathrm{PAI}}_{ei}{W}_i$

${\mathrm{WAI}}_e=2\underset{i=1}{\overset{9}{\Σ}}{\mathrm{WAI}}_{ei}{W}_i$

In formula, W_iFor the weight that i-th zenith angle is interval, it is respectively 0.0038 in weight that 9 zenith angles are interval, 0.0303、0.0597、0.0873、0.1122、0.1337、0.1512、0.1640、0.2578。

Step S513: the effective wooden area index WAI of sampled point_eiWith gross area indices P AI_eiRatio be sampled point The wooden gross area of Vegetation canopy compares parameter.

Further, in described step S5, the calculating of described aggregate index is based on classification image, wooden by extracting Component, canopy solvent clearance rate or wood components and canopy solvent line-transect, according to canopy solvent and wooden group Segregation aggregate index number algorithm calculates Vegetation canopy wood components and canopy solvent aggregate index, and concrete calculation procedure is as follows:

Step S521: view zenith angle is divided into 81 zenith angle intervals, step-length is 1 °, is i.e. divided into 0 °-80 °；One by one All azimuth directions classification image circulation interval to each zenith angle, sky, leaves, wood components and soil group on cumulative image The number of pixels divided, obtains each zenith angle direction sky, leaves, wood components and soil constitution sum of all pixels；

Step S522: each zenith angle interval sky, leaves and the pixel summation of soil constitution and the ratio of image pixel sum Value is the clearance rate p (θ of wood components_wi), and the ratio of the pixel summation of sky and image pixel sum to be canopy basic Clearance rate p (the θ of component_pi), canopy solvent CI that each zenith angle is interval_e(θ_i) and wood components aggregate index CI_w(θ_i) meter Calculation method is as follows:

${\mathrm{CI}}_e\left({\mathrm{\θ}}_i\right)=\frac{ln\[\stackrel{\‾}{p\left({\mathrm{\θ}}_{pi}\right)}\]}{\stackrel{\‾}{\ln \[p\left({\mathrm{\θ}}_{pi}\right)\]}}$

${\mathrm{CI}}_w\left({\mathrm{\θ}}_i\right)=\frac{ln\[\stackrel{\‾}{p\left({\mathrm{\θ}}_{wi}\right)}\]}{\stackrel{\‾}{\ln \[p\left({\mathrm{\θ}}_{wi}\right)\]}}.$

Further, in described step S5, the calculating of described sun photosynthetically active radiation distribution is with half direction of bowl classification shadow Based on the clearance rate of wood components, leaves and sky component, calculate according to sun photosynthetically active radiation distributed computing model The each zenith angle of Vegetation canopy and the distribution of azimuth direction sun photosynthetically active radiation, concrete calculation procedure is as follows:

Step S531: in zenith angle and azimuth direction, hemisphere image is divided into 9 and 12 deciles, i.e. subdivision respectively is 108 annulus, add up sky, leaves, wood components and soil constitution sum of all pixels on each zenith angle annulus one by one, To each zenith angle annulus sky, leaves, wood components and soil constitution sum of all pixels；

Step S532: in each zenith angle annulus, sky component pixel summation is each zenith with the ratio of annulus sum of all pixels Clearance rate p (the θ of angle annulus canopy solvent_{m_n}), the pixel summation of all annulus skies and this zenith in each zenith angle interval The ratio of angle range annulus sum of all pixels is this zenith angle interval canopy solvent clearance rate p (θ_i)；If sky is homogeneous Sky, Vegetation canopy top sun direct projection PAR radiant intensity is 0.55, and sky scattering PAR radiant intensity is 0.45, then sampled point Sun direct projection PAR (PAR_Dir) and sky scattering PAR (PAR_Dif) computational methods are as follows:

PAR_Dir=0.55*p (θ_{m_n})

PAR_Dif=0.45*p (θ_i)

Wherein, p (θ_{m_n}) it is sun incidence zenith angle and the canopy solvent gap in azimuth direction place annulus interval Rate.

Compared with prior art, present invention have the advantage that a set of low cost of offer, multipurpose, extensibility are strong The implementation method of Vegetation canopy multi-optical spectrum imaging system, can obtain Vegetation canopy multispectral image based on this system and method, and The wooden gross area of Vegetation canopy can be extracted based on Vegetation canopy multispectral image to divide than parameter, aggregate index and photosynthetically active radiation Cloth, thus improve Vegetation canopy structural parameters and photosynthetically active radiation distribution ground survey precision.

Accompanying drawing explanation

Fig. 1 is the implementing procedure figure of the embodiment of the present invention.

Fig. 2 is visible light wave range Vegetation canopy multispectral image schematic diagram in the embodiment of the present invention.

Fig. 3 is near infrared band Vegetation canopy multispectral image schematic diagram in the embodiment of the present invention.

Fig. 4 is the canopy constituent classification image schematic diagram of Vegetation canopy multispectral image in the embodiment of the present invention.

Detailed description of the invention

Below in conjunction with the accompanying drawings and embodiment the present invention will be further described.

The present embodiment provides the implementation method of a kind of Vegetation canopy multi-optical spectrum imaging system, as it is shown in figure 1, include following step Rapid:

Step S1: transformation slr camera Nikon D300S makes it possess near infrared band imaging capability；

Step S2: camera is placed in The Cloud Terrace (Man Futu pan and tilt head MH303SPH) and spider (Man Futu MT055CXPRO4), rotary platform, to setting zenith angle and azimuth direction, installs fish eye lens (Sigma 4.5mm flake mirror Head) or telephoto lens (Nikon Nikon AF-SDX 18-200), and (Nature HL-OPTICS cuts red to install filter on camera lens Outer filter), Nature HL-OPTICS 760nm, Nature HL-OPTICS 850nm)；

Step S3: make camera in the single wave band of near-infrared or visible ray and two wave band imagings of near-infrared, according to setting Scheme completes the canopy multi-spectra imaging of episphere direction；

When camera lens is fish eye lens (Sigma 4.5mm fish eye lens), adjusts The Cloud Terrace and make camera lens towards extremely Required direction, such as zenith direction, is perpendicular to ground level etc., arranges camera imaging parameter, uses shutter control camera respectively closely Infrared and visible light wave range imaging.

When camera lens is telephoto lens (Nikon Nikon AF-S DX 18-200), Vegetation canopy multispectral image is observed Step is as follows:

1.. adjust The Cloud Terrace and make camera lens towards to required direction, such as zenith direction, be perpendicular to ground level etc., phase is set Machine imaging parameters, installs in telephoto lens and cuts infrared filter, use shutter control camera visible light wave range imaging (380nm- 700nm)。

2.. the azimuth of holding The Cloud Terrace and zenith angle, towards constant, installed the infrared filter of 760nm in telephoto lens, are made With shutter control camera near infrared band (760nm-960nm) imaging.

3.. the azimuth of holding The Cloud Terrace and zenith angle, towards constant, installed the infrared filter of 850nm in telephoto lens, are made With shutter control camera near infrared band (850nm-960nm) imaging.

4.. adjust The Cloud Terrace to required azimuth and zenith angle direction, repeat 1.-3. step, until complete to set is all Azimuth and the observation of zenith angle direction.

Step S4: use ENVI 4.7 remote sensing to process software and Vegetation canopy multispectral image is classified, its step of classifying Rapid as follows:

1.. use the method for artificial coupling to same azimuth and the visible ray in zenith angle direction and near-infrared picture to carrying out Registration, the method choice " Image-to-Image " of Image Matching, the number of control points of coupling can not be less than 10, and control point Need to be uniformly distributed in imagery zone, control point relative error should be less than 1.5 pixels.

2.. by the picture matched to using the classification of ISODATA not supervised classification, the parameter of point supervised classification is arranged For: subclass numbers is 24, and standard variance is 0.5, and cycle-index can be set to 50-60, and every class minimum pixel number can be set to 30.Enter 24 subclasses are sorted out to four big classes such as sky, leaves, wood components and soil by step respectively, the subclass of each big apoplexy due to endogenous wind are closed And, the classification video conversion after merging is bmp format image.

Step S5: calculate the wooden gross area of Vegetation canopy based on image of classifying than parameter, aggregate index and sunlight Conjunction Net long wave radiation is distributed.

In the present embodiment, in described step S5, the wooden gross area of described Vegetation canopy than the technology of parameter with shadow of classifying Based on the clearance rate of upper wood components, leaves and sky component, according to Beer law computational methods calculate sampled point PAI and WAI, and then derive the wooden gross area of sampled point Vegetation canopy than parameter, concrete calculation procedure is as follows:

Step S511: zenith angle direction is in turn divided into 9 zenith angle intervals, i.e. (0 °-5 °), (5 °-15 °), (15 °- 25 °), (25 °-35 °), (35 °-45 °), (45 °-55 °), (55 °-65 °), (> 65 °-75 °), (75 °-85 °), statistics is each respectively Sky, leaves, wood components and the number of pixels of soil constitution on individual classification image, each azimuth side interval to each zenith angle On classification image, sky, leaves, the number of pixels of wood components and soil constitution add up, obtain 9 zenith angle direction skies, The sum of all pixels of leaves, wood components and soil constitution；

Step S512: each zenith angle interval sky, leaves and the pixel summation of soil constitution and the ratio of image pixel sum Value is the clearance rate p (θ of wood components_wi), and the ratio of the pixel summation of sky and image pixel sum to be canopy basic Clearance rate p (the θ of component_pi), effective gross area indices P AI that each zenith angle is interval_eiWith wooden area index WAI_eiComputational methods As follows:

PAI_ei=-ln [p (θ_pi)]cos(θ_i), W_i=sin (θ_i)d(θ_i)

WAI_ei=-ln [p (θ_wi)]cos(θ_i), W_i=sin (θ_i)d(θ_i)

In formula, i is zenith angle Interval Coding, θ_iFor zenith angle；

Sampled point effective gross area indices P AI_eWith wooden area index WAI_eComputational methods are as follows:

${\mathrm{PAI}}_e=2\underset{i=1}{\overset{9}{\Σ}}{\mathrm{PAI}}_{ei}{W}_i$

${\mathrm{WAI}}_e=2\underset{i=1}{\overset{9}{\Σ}}{\mathrm{WAI}}_{ei}{W}_i$

In formula, W_iFor the weight that i-th zenith angle is interval, it is respectively 0.0038 in weight that 9 zenith angles are interval, 0.0303、0.0597、0.0873、0.1122、0.1337、0.1512、0.1640、0.2578。

Step S513: the effective wooden area index WAI of sampled point_eiWith gross area indices P AI_eiRatio be sampled point The wooden gross area of Vegetation canopy compares parameter.

In the present embodiment, in described step S5, the calculating of described aggregate index is based on classification image, by extracting Wood components, canopy solvent clearance rate or wood components and canopy solvent line-transect, according to canopy solvent and wood Matter component aggregate index algorithm calculates Vegetation canopy wood components and canopy solvent aggregate index, and concrete calculation procedure is such as Under:

Step S521: view zenith angle is divided into 81 zenith angle intervals, step-length is 1 °, is i.e. divided into 0 °-80 °；One by one All azimuth directions classification image circulation interval to each zenith angle, sky, leaves, wood components and soil group on cumulative image The number of pixels divided, obtains each zenith angle direction sky, leaves, wood components and soil constitution sum of all pixels；

Step S522: each zenith angle interval sky, leaves and the pixel summation of soil constitution and the ratio of image pixel sum Value is the clearance rate p (θ of wood components_wi), and the ratio of the pixel summation of sky and image pixel sum to be canopy basic Clearance rate p (the θ of component_pi), canopy solvent CI that each zenith angle is interval_e(θ_i) and wood components aggregate index CI_w(θ_i) meter Calculation method is as follows:

${\mathrm{CI}}_e\left({\mathrm{\θ}}_i\right)=\frac{ln\[\stackrel{\‾}{p\left({\mathrm{\θ}}_{pi}\right)}\]}{\stackrel{\‾}{\ln \[p\left({\mathrm{\θ}}_{pi}\right)\]}}$

${\mathrm{CI}}_w\left({\mathrm{\θ}}_i\right)=\frac{ln\[\stackrel{\‾}{p\left({\mathrm{\θ}}_{wi}\right)}\]}{\stackrel{\‾}{\ln \[p\left({\mathrm{\θ}}_{wi}\right)\]}}.$

In the present embodiment, in described step S5, the calculating of described sun photosynthetically active radiation distribution divides with half direction of bowl Based on the clearance rate of class image wood components, leaves and sky component, according to sun photosynthetically active radiation distributed computing model Calculating each zenith angle of Vegetation canopy and the distribution of azimuth direction sun photosynthetically active radiation, concrete calculation procedure is as follows:

Step S531: in zenith angle and azimuth direction, hemisphere image is divided into 9 and 12 deciles, i.e. subdivision respectively is 108 annulus, add up sky, leaves, wood components and soil constitution sum of all pixels on each zenith angle annulus one by one, To each zenith angle annulus sky, leaves, wood components and soil constitution sum of all pixels；

Step S532: in each zenith angle annulus, sky component pixel summation is each zenith with the ratio of annulus sum of all pixels Clearance rate p (the θ of angle annulus canopy solvent_{m_n}), the pixel summation of all annulus skies and this zenith in each zenith angle interval The ratio of angle range annulus sum of all pixels is this zenith angle interval canopy solvent clearance rate p (θ_i)；If sky is homogeneous Sky, Vegetation canopy top sun direct projection PAR radiant intensity is 0.55, and sky scattering PAR radiant intensity is 0.45, then sampled point Sun direct projection PAR (PAR_Dir) and sky scattering PAR (PAR_Dif) computational methods are as follows:

PAR_Dir=0.55*p (θ_{m_n})

PAR_Dif=0.45*p (θ_i)

Wherein, p (θ_{m_n}) it is sun incidence zenith angle and the canopy solvent gap in azimuth direction place annulus interval Rate.

Use the Vegetation canopy multispectral image that obtains of above method and canopy constituent classification schematic diagram such as Fig. 2-Fig. 4 institute thereof Show.

Examples detailed above employing finite length average algorithm is as Vegetation canopy aggregate index Measurement Algorithm, but the present invention is adopted Vegetation canopy aggregate index Measurement Algorithm do not limited by examples detailed above, as gap length Distribution Algorithm, combination method equivalent Sample may be used without the canopy multi-spectra radiographic measurement Vegetation canopy aggregate index related in the present invention, other any without departing from this The change made under spirit and the principle of invention, modify, substitute, combine, simplify, all should be equivalent substitute mode, all Within being included in protection scope of the present invention.

When examples detailed above carries out the wooden gross area than parameter measurement, effective gross area index and effective wooden area index ginseng Number do not consider that canopy building-up effect affects, but the wooden gross area of the present invention than measurement method of parameters not by examples detailed above Restriction, it participates in wooden gross area gross area index and wooden area index than parameter measurement it is contemplated that canopy building-up effect Impact, the change made, modifies, substitutes, combines, simplifies under other any spirit without departing from the present invention and principle, All should be the substitute mode of equivalence, within being included in protection scope of the present invention.

It is homogeneous sky that examples detailed above carries out hypothesis sky when the distribution of Vegetation canopy sun photosynthetically active radiation calculates, vegetation Canopy top sun direct projection PAR radiant intensity is 0.55, and sky scattering PAR radiant intensity is 0.45, but of the present invention The distribution of Vegetation canopy sun photosynthetically active radiation calculates and is not limited by above-mentioned assumed condition, other any without departing from the present invention Spirit and principle under made change, modify, substitute, combine, simplify, all should be the substitute mode of equivalence, all comprise Within protection scope of the present invention.The foregoing is only presently preferred embodiments of the present invention, all according to scope of the present invention patent Impartial change and the modification done, all should belong to the covering scope of the present invention.

Claims (10)

1. the implementation method of a Vegetation canopy multi-optical spectrum imaging system, it is characterised in that: comprise the following steps:

Step S1: transform a slr camera so that it is near infrared band imaging；

Step S2: be placed on The Cloud Terrace and spider by described slr camera, rotary platform is to setting zenith angle and azimuth side To, fish eye lens, telephoto lens or tight shot are installed, and filter is installed on camera lens；

Step S3: make described slr camera become at the single wave band of near-infrared or visible ray and two wave bands of near-infrared or multiple wave band Picture, completes episphere direction Vegetation canopy multispectral imaging；

Step S4: the Vegetation canopy multispectral image of collection is classified；

Step S5: calculate the wooden gross area of Vegetation canopy based on image of classifying more photosynthetic than parameter, aggregate index and the sun Net long wave radiation is distributed.

The implementation method of a kind of Vegetation canopy multi-optical spectrum imaging system the most according to claim 1, it is characterised in that: described In step S1, described slr camera include Nikon D300, Nikon D80, Canon 5D, Canon 70D, Canon 700D, Nikon D7100 and micro-one camera.

A kind of Vegetation canopy multi-optical spectrum imaging system the most according to claim 1 and method, it is characterised in that: described step In S1, transforming the machine that described list is anti-phase is that the IR-cut filter by anterior for internal for slr camera CCD original paper replaces with Near-infrared filter, described IR-cut filter is low passage filter.

A kind of Vegetation canopy multi-optical spectrum imaging system the most according to claim 1 and method, it is characterised in that: described step In S2, described The Cloud Terrace realizes 0 ° of-360 ° of direction at azimuth direction and rotates, and realizes in zenith angle direction turning in 0 ° of-180 ° of direction Dynamic.

A kind of Vegetation canopy multi-optical spectrum imaging system the most according to claim 1 and method, it is characterised in that: described step In S3, described visible ray and two wave band imagings of near-infrared realize drift angle and side on the same day by the method for filter disc before replacing camera lens Parallactic angle direction Vegetation canopy multispectral imaging.

A kind of Vegetation canopy multi-optical spectrum imaging system the most according to claim 1 and method, it is characterised in that: described step In S3, when the camera lens that slr camera carries is fish eye lens, described slr camera lens direction should be towards zenith direction, described list Anti-phase machine single imaging can realize episphere direction Vegetation canopy multispectral imaging；The camera lens carried when slr camera camera is When focal length or tight shot, half direction of bowl is divided into several zenith angles and azimuth respectively in zenith angle and azimuth direction Interval, described slr camera all carries out Vegetation canopy multispectral imaging in each zenith angle and azimuth direction.

A kind of Vegetation canopy multi-optical spectrum imaging system the most according to claim 1 and method, it is characterised in that: described step In S4, the classification of described Vegetation canopy multispectral image use supervised classification or point supervised classification method to Vegetation canopy visible ray and Near infrared band image classification, then Vegetation canopy image is divided into wood components, leaves and sky three major types.

A kind of Vegetation canopy multi-optical spectrum imaging system the most according to claim 1 and method, it is characterised in that: described step In S5, the wooden gross area of described Vegetation canopy than the technology of parameter with wood components on classification image, leaves and sky component Based on clearance rate, calculate sampled point PAI and WAI according to Beer law computational methods, and then derive sampled point Vegetation canopy The wooden gross area is than parameter, and concrete calculation procedure is as follows:

Step S511: zenith angle direction is in turn divided into 9 zenith angle intervals, i.e. (0 °-5 °), (5 °-15 °), (15 °- 25 °), (25 °-35 °), (35 °-45 °), (45 °-55 °), (55 °-65 °), (> 65 °-75 °), (75 °-85 °), statistics is each respectively Sky, leaves, wood components and the number of pixels of soil constitution on individual classification image, each azimuth side interval to each zenith angle On classification image, sky, leaves, the number of pixels of wood components and soil constitution add up, obtain 9 zenith angle direction skies, The sum of all pixels of leaves, wood components and soil constitution；

Step S512: the pixel summation of each zenith angle interval sky, leaves and soil constitution with the ratio of image pixel sum is Clearance rate p (θ for wood components_wi), and the ratio of the pixel summation of sky and image pixel sum is canopy solvent Clearance rate p (θ_pi), effective gross area indices P AI that each zenith angle is interval_eiWith wooden area index WAI_eiComputational methods are such as Under:

PAI_ei=-ln [p (θ_pi)]cos(θ_i), W_i=sin (θ_i)d(θ_i)

WAI_ei=-ln [p (θ_wi)]cos(θ_i), W_i=sin (θ_i)d(θ_i)

In formula, i is zenith angle Interval Coding, θ_iFor zenith angle；

Sampled point effective gross area indices P AI_eWith wooden area index WAI_eComputational methods are as follows:

${\mathrm{PAI}}_e=2\underset{i=1}{\overset{9}{\Σ}}{\mathrm{PAI}}_{ei}{W}_i$

${\mathrm{WAI}}_e=2\underset{i=1}{\overset{9}{\Σ}}{\mathrm{WAI}}_{ei}{W}_i$

In formula, W_iFor the weight that i-th zenith angle is interval, it is respectively 0.0038 in weight that 9 zenith angles are interval, 0.0303, 0.0597、0.0873、0.1122、0.1337、0.1512、0.1640、0.2578。

Step S513: the effective wooden area index WAI of sampled point_eiWith gross area indices P AI_eiRatio be sampled point vegetation hat The wooden gross area of layer compares parameter.

A kind of Vegetation canopy multi-optical spectrum imaging system the most according to claim 1 and method, it is characterised in that: described step In S5, the calculating of described aggregate index by classification image based on, by extract wood components, canopy solvent clearance rate or Wood components and canopy solvent line-transect, calculate Vegetation canopy according to canopy solvent and wood components aggregate index algorithm Wood components and canopy solvent aggregate index, concrete calculation procedure is as follows:

Step S521: view zenith angle is divided into 81 zenith angle intervals, step-length is 1 °, is i.e. divided into 0 °-80 °；One by one to respectively Zenith angle interval all azimuth directions classification image circulation, sky, leaves, wood components and soil constitution on cumulative image Number of pixels, obtains each zenith angle direction sky, leaves, wood components and soil constitution sum of all pixels；

Step S522: the pixel summation of each zenith angle interval sky, leaves and soil constitution with the ratio of image pixel sum is Clearance rate p (θ for wood components_wi), and the ratio of the pixel summation of sky and image pixel sum is canopy solvent Clearance rate p (θ_pi), canopy solvent CI that each zenith angle is interval_e(θ_i) and wood components aggregate index CI_w(θ_i) calculating side Method is as follows:

${\mathrm{CI}}_e\left({\mathrm{\θ}}_i\right)=\frac{ln\[\stackrel{\‾}{p\left({\mathrm{\θ}}_{pi}\right)}\]}{\stackrel{\‾}{ln\[p\left({\mathrm{\θ}}_{pi}\right)\]}}$

${\mathrm{CI}}_w\left({\mathrm{\θ}}_i\right)=\frac{ln\[\stackrel{\‾}{p\left({\mathrm{\θ}}_{wi}\right)}\]}{\stackrel{\‾}{ln\[p\left({\mathrm{\θ}}_{wi}\right)\]}}.$

A kind of Vegetation canopy multi-optical spectrum imaging system the most according to claim 1 and method, it is characterised in that: described step In rapid S5, the calculating of described sun photosynthetically active radiation distribution is with half direction of bowl classification image wood components, leaves and sky group Based on the clearance rate divided, calculate each zenith angle of Vegetation canopy and azimuth according to sun photosynthetically active radiation distributed computing model Direction sun photosynthetically active radiation distribution, concrete calculation procedure is as follows:

Step S531: in zenith angle and azimuth direction, hemisphere image is divided into 9 and 12 deciles, i.e. subdivision respectively is 108 Annulus, adds up sky, leaves, wood components and soil constitution sum of all pixels on each zenith angle annulus one by one, obtains each Zenith angle annulus sky, leaves, wood components and soil constitution sum of all pixels；

Step S532: in each zenith angle annulus, sky component pixel summation is each zenith angle circle with the ratio of annulus sum of all pixels Clearance rate p (the θ of garlands layer solvent_{m_n}), the pixel summation of all annulus skies and this zenith angular region in each zenith angle interval Between the ratio of annulus sum of all pixels be this zenith angle interval canopy solvent clearance rate p (θ_i)；Assume that sky is homogeneous sky Sky, Vegetation canopy top sun direct projection PAR radiant intensity is 0.55, and sky scattering PAR radiant intensity is 0.45, then sampled point is too Sun direct projection PAR (PAR_Dir) and sky scattering PAR (PAR_Dif) computational methods are as follows:

PAR_Dir=0.55*p (θ_{m_n})

PAR_Dif=0.45*p (θ_i)

Wherein, p (θ_{m_n}) it is sun incidence zenith angle and the canopy solvent clearance rate in azimuth direction place annulus interval.