CN101319896B - Method for measuring three-dimensional structure of forest gap - Google Patents

Abstract

The invention relates to a method for measuring a solid structure of a forest gap which includes the following steps of: using a digital camera provided with a fish-eye lens to vertically upwards shoot two semi-spherical images at the different height of any position in the forest gap; measuring the parameters of the forest gap; processing the two semi-spherical images and recording the coordinates of the edge points of a forest canopy in the forest gap; calculating the area and the shape of the forest gap and the heights of the forest canopy of the forest gap of each orientation according to the protection principle of the fish-eye lens as well as each coordinate point and the parameters of the forest gap. The invention firstly provides a method for measuring the solid structure of the forest gap, can fast and objectively measure the area and the shape of the forest gap and the heights of the edge wood of the forest gap of each orientation, is easily used, only two semi-spherical images is shooted when measuring the area and the shape of the forest gap, only two parameters of a gradient and an aspect are added when calculating the heights of the edge wood of the forest gap of each orientation with higher measuring precision; besides, the measuring result is not affected by the measuring position.

Description

A kind of method of measuring three-dimensional structure of forest gap

Technical field

The present invention relates in forestry and the ecology in the field about the measuring method of woods window feature specifically a kind of method of measuring three-dimensional structure of forest gap.

Background technology

Woods window (gap) refers to will being occupied and the space of upgrading by new individuality of forming by the above storey trees death of a strain, and this notion is followed the research of forest cycle and produced.It is that recurrent important small scale is disturbed in the forest community that the woods window disturbs, and has become one of current forest ecology research most active fields.Only change the most violent envirment factor in the woods window, owing to be subjected to the three-dimensional structure of forest gap (mainly referring to size, shape, different azimuth woods window edge wood canopy height) of sun altitude and azimuthal diurnal variation and seasonal variety, complexity and the influence of woods window mima type microrelief, the illumination of woods window not only exists diurnal variation and seasonal variations, also shows complicated three dimensions heterogeneity simultaneously.Simultaneously, the change in time and space of illumination also causes thermodynamic features generation respective change such as surface temperature, air temperature/humidity in the woods window, and photoenvironment and thermodynamic features can influence envirment factors such as soil moisture, physicochemical property, nutrient decomposition, soil microbial activities, thereby cause the heterogeneity of woods window envirment factor.Since light in the woods window in uneven distribution and the woods window directly incident light have Time of Day and seasonal variation, so it is very difficult to measure in the woods window luminous environment.Because sun altitude and azimuthal variation, the illumination difference that diverse location receives in the woods window is very big, so direct measuring method (as illuminometer and light quantum probe) all can only monitor the variation of intensity of illumination in monitoring time of limited quantity monitoring point in the woods window.The method of some other indirect measuring light environment (fish eye lens method) once also can only be measured certain any photoenvironment though have advantages such as simple, quick.In addition, these methods only can be estimated woods window illumination variation at grade.Yet the blade of plant lays respectively at different height layers, therefore, measures the photoenvironment that could obtain plant in the woods window apart from the illumination heterogeneity of ground differing heights layer exactly.(gap light index GLI) can estimate the illumination condition of woods window optional position to the woods window optical index that Canham proposes.Three-dimensional structure of forest gap is the key factor that influences illumination in the woods window, therefore, is the key parameter of GLI.GLI is reduced to vertical cylindrical body to the spatial structure of woods window.Three-dimensional structure of forest gap mainly comprises the height of woods window size, shape and each orientation woods window edge wood canopy.Though the existing many methods of measuring woods window sizes so far, do not have a kind of method of simple, objective, general measurement woods window size; Simultaneously, few to the measurement Research of woods window shape, many researchers are divided into linear and non-linear two classes to the woods window according to the ratio of woods window major axis and minor axis simply; In addition, also there is not research to relate to the measuring method of woods window edge wood at present at the canopy height of different azimuth.Therefore, find a kind of quick, objective, accurately to measure the method for three-dimensional structure of forest gap significant for the research of woods window illumination.

Summary of the invention

For remedying the technological gap of three-dimensional structure of forest gap measuring method in the prior art, the object of the present invention is to provide a kind of can be objective, accurate and the method for easy-operating measurement three-dimensional structure of forest gap.

For achieving the above object, the technical solution used in the present invention is:

The differing heights of any one position is taken two hemisphere face images vertically upward with the fish-eye digital camera of assembling in the woods window; Measure woods window parameter; Two hemisphere face images are handled the coordinate of record photograph middle forest window crown canopy marginal point; According to the height of fish-eye projection theory with area, shape and each orientation woods window crown canopy of each coordinate points and woods window calculation of parameter woods window.

The photographing request of wherein said hemisphere face image is: select optional position in the woods window of calm, full cloudy day, will be equipped with fish-eye digital camera and be placed on the tripod, take 2 hemisphere face images vertically upward at differing heights; Camera aperture transfers to minimum during shooting, and focal length transfers to the infinite distance, guarantee the hemisphere face image directly over for the compass direct north; Described employing image processing software is handled the hemisphere face image, is that central point with the hemisphere face image is that the coordinate of crown canopy marginal point in the photograph is write down at interval since 0 degree position angle and with fixed angle in the center, and every sheet photo writes down n=360/g coordinate points altogether; Described woods window parameter comprises the gradient and aspect.

The woods window is approximated to n limit shape, calculates forest gap area by following formula:

$A=0.5\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{D}_{i\·g}\·D_{(i-1)\·g}\·\sin (2\mathrm{\π}/n)$

Wherein, ig is the position angle of i coordinate points, D _{I g}Be the vertical range of ig orientation crown canopy marginal point to the fish eye lens optical axis.

The woods window is approximated to n limit shape, and the woods window shape is calculated by shape index SI:

$\mathrm{SI}=P/2\sqrt{\mathrm{\πA}}$

P is a woods window girth in the formula, and A is a forest gap area.

Described woods window girth P calculates by following formula:

$P=\underset{i=1}{\overset{n}{\mathrm{\Σ}}}\sqrt{{(D_{i\·g}-D_{(i-1)g}\·\cos (2\mathrm{\π}/n))}^2+{(D_{(i-1)\·g}\·\sin (2\mathrm{\π}/n))}^2}$

Wherein, ig is the position angle of i coordinate points, D _IgBe the vertical range of ig orientation crown canopy marginal point to the fish eye lens optical axis.

The woods window is approximated to n limit shape, and each orientation crown canopy marginal point is apart from the height H on ground _{T, α}Calculate by following formula:

H _t，α＝H _α+D _α·cos(α-β)·tan(γ)

Wherein, α, β, γ are respectively position angle, aspect, the gradient arbitrarily, H _α, D _αBe respectively α orientation crown canopy marginal point above camera lens height and to the vertical range of fish eye lens optical axis.

The height H of α orientation crown canopy marginal point above camera lens _αWith vertical range D to the fish eye lens optical axis _αComputing method are as follows:

H _α＝d·tan(θ _h，α)/(tan(θ _h，α)-tan(θ _l，α))+h

D _α＝d·tan(θ _h，α)·tan(θ _l，α)/(tan(θ _h，α)-tan(θ _l，α))

In the formula, d, h be respectively the discrepancy in elevation of camera lens when taking for twice, camera lens height overhead when lower is taken, θ _{L, α}, θ _{H, α}α orientation crown canopy marginal point is to the zenith angle of fish eye lens vertical optical axis when being respectively the shooting of lower and eminence.

Described zenith angle obtains by the following method:

In two hemisphere face images, write down the coordinate of each orientation crown canopy marginal point respectively, according to coordinate figure calculate the azimuth angle alpha of each orientation crown canopy marginal point and subpoint thereof to photo center point apart from r, ask each orientation porch to put the zenith angle θ of fish eye lens vertical optical axis according to fish-eye polar coordinates projection theory formula:

θ/90＝r/R

In the formula, R is the radius distance of θ=90 o'clock subpoint in the hemisphere face image.

The present invention has following beneficial effect and advantage:

1. propose first to measure the three-dimensional structure of forest gap method, can measure the height of forest gap area, shape and each orientation woods window edge wood fast, objectively;

2. use simply, only need take 2 hemisphere face images when measuring woods window size and shape, calculate each orientation woods window edge wood and only need add the gradient and 2 parameters of aspect highly the time;

3. has higher measuring accuracy;

4. measurement result is not subjected to the influence of measuring position.

Description of drawings

Fig. 1 is the polar coordinates projection theory figure of hemisphere face image;

Fig. 2 A measures the schematic diagram of woods window size and shape for the two hemisphere face image methods of the present invention;

Fig. 2 B measures woods window edge wood elevation principle figure for the two hemisphere face image methods of the present invention;

Fig. 3 A is the two hemisphere face image method data processing synoptic diagram of the present invention;

Fig. 3 B is the partial enlarged drawing of Fig. 3 A;

Fig. 4 is the instance graph that these two hemisphere face image methods calculate gained.

Embodiment

(1) experiment place

Sample plot click take from certain forest ecology experiment centre (41 ° of 51.102 ' N, 124 ° 54.543 ' E), height above sea level 252～1116m, weather belong to the continental monsoon climate in warm temperate zone, winter very long cold, summer is burning hot and rainy.3.9～5.4 ℃ of average temperatures of the whole year, 36.5 ℃ of the extreme highest temperatures, minimum subzero 37.6 ℃.Greater than 2497.5～2943.0 ℃ of 10 ℃ year active accumulated temperatures, frostless season 120～139d, annual 2433h at sunshine, annual precipitation 700～850mm, rainfall concentrates on for 6～August.Vegetation is under the jurisdiction of the Changbai Mountain fauna, based on Natural Secondary Forests, constructive species has Fraxinus rhynchophylla (Fraxinus rhynchophylla), Mongolian oak (Quercus mongolica), look wood maple (Acer mono), juglans mandshurica (Juglansmandshurica) etc., and part larix olgensis (Larix olgensis) artificial forest and Korean pine (Pinuskoraiensis) are arranged.

(2) field survey

Selecting 12 artificial forest windows that vary in size is the respondent, select calm, the full cloudy day, settle tripod at each artificial forest window center point and 2 random points, positioned vertical is furnished with fish-eye digital camera, 2 differing heights are respectively taken 1 hemisphere face image in each position, 2 times are taken the discrepancy in elevation is 0.5m, and lower shooting point fish eye lens is than the high 1m in ground; Camera aperture transfers to minimum during shooting, and focus transfers to the infinite distance, simultaneously, selects 5 strain woods window edges wood to measure average height H in each woods window at random, and measures gradient γ and aspect β.In addition, settle forestry compass at each woods window center point, measure along 16 compass headings (0,22.5,45 ..., 337.5 degree) to the domatic distance that is parallel to of crown canopy edge-perpendicular subpoint.

(3) calculating of forest gap area

(1) use of two hemisphere face image methods

Select full the moon and calm sky optional position in the woods window will be equipped with fish-eye digital camera positioned vertical on tripod; Camera aperture transfers to minimum, and focus transfers to the infinite distance; Take 2 hemisphere face images vertically upward, take for 2 times at woods window same position but a difference in height h is arranged, and guarantee the hemisphere face image directly over be the compass direct north.Then photo is imported in the computer, handled with image processing software (as, Adobe Illustrator): write down the coordinate of photograph middle forest window crown canopy marginal point since 0 degree at interval with the g degree, every sheet photo writes down n=360/g coordinate points altogether.The height of size, shape and the n orientation woods window edge wood of last woods window can calculate according to the method for introducing below.

(2) two hemisphere face image ratio juris

The imaging mode of hemisphere face image is polar coordinates projection pattern (as shown in Figure 1), that is, woods window crown canopy marginal point to the zenith angle θ of fish eye lens vertical optical axis and subpoint to photo center point apart from r linear (formula 1).

θ/90＝r/R (1)

In the formula, R is the radius distance of θ=90 o'clock subpoint in the hemisphere face image.

For the coordinate of each orientation α crown canopy marginal point in two sheet photos, can try to achieve r, can try to achieve the zenith angle θ of α orientation crown canopy marginal point in the hemisphere face image that lower, eminence are taken respectively according to formula (1) then _{L, α}, θ _{H, α}

According to geometric relationship (shown in Fig. 2 A), can obtain formula 2,3:

H _α＝d·tan(θ _h，α)/(tan(θ _h，α)-tan(θ _l，α))+h (2)

D _α＝d·tan(θ _h，α)·tan(θ _l，α)/(tan(θ _h，α)-tan(θ _l，α)) (3)

In the formula, D _αBe the vertical range of α orientation crown canopy marginal point to the fish eye lens optical axis, H _αBe the height of α orientation crown canopy marginal point above fish eye lens, d, h be respectively the discrepancy in elevation of camera lens when taking for twice, camera lens height overhead when lower is taken, θ _{L, α}, θ _{H, α}α orientation crown canopy marginal point is to the zenith angle of fish eye lens vertical optical axis when being respectively the shooting of lower and eminence.

Crown canopy woods window is approximately n limit polygon, and then the area A of woods window and girth P can try to achieve according to formula (4) and (5),

$A=0.5\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{D}_{i\·g}\·D_{(i-1)\·g}\·\sin (2\mathrm{\π}/n)---\left(4\right)$

$P=\underset{i=1}{\overset{n}{\mathrm{\Σ}}}\sqrt{{(D_{i\·g}-D_{(i-1)g}\·\cos (2\mathrm{\π}/n))}^2+{(D_{(i-1)g}\·\sin (2\mathrm{\π}/n))}^2}---\left(5\right)$

Adopt shape index (SI) to calculate the woods window shape,

$\mathrm{SI}=P/2\sqrt{\mathrm{\πA}}---\left(6\right)$

Do not consider the influence (being that the domatic gradient and aspect is definite value) of mima type microrelief, α orientation crown canopy marginal point is apart from the height H on ground _{T, α}Can try to achieve according to formula (7) (seeing Fig. 2 B):

H _t，α＝H _α+D _α·cos(α-β)·tan(γ) (7)

In the formula, β, γ are respectively the aspect and the gradient.

H _e，γ＝H-tan(α)·D _γ·cos(γ-β)-h (8)

Above-mentioned various in, h is the height of camera lens apart from ground, ig is the position angle of i coordinate points, D _IgBe the vertical range of ig orientation crown canopy marginal point to the fish eye lens optical axis, last some i=n+1 also is starting point i=1.

(3) isogonism 16 limit shape methods (equiangular sixteen-sided polygon method, ESM) (according to example):

Being parallel to domatic distance transform is horizontal range, and note is done: l ₁, l ₂..., l ₁₆, l ₁Compass heading be 0 degree, the position angle of other l increases successively with 22.5 degree, to l ₁₆Compass heading is 337.5 degree.The computing formula of equiangular polygon method is specially:

$A=0.5\underset{i=1}{\overset{16}{\mathrm{\Σ}}}{l}_{i+1}\·l_i\·\sin (2\mathrm{\π}/n)$

Wherein, during i=n, l _I+1Equal l ₁

(4) data processing

Shown in Fig. 3 A, 3B, in image processing software Illustrator, the hemisphere face image is amplified to 1600%.Spending and spend with 10 since 0 in the hemisphere face image serves as to write down woods window crown canopy marginal point coordinate at interval, calculate the r of each orientation crown canopy marginal point according to coordinate figure, the last height that calculates size, shape and 36 orientation woods window crown canopies of each woods window according to two hemisphere face image method principles.An example of the three-dimensional structure of forest gap that Fig. 4 records for two hemisphere face image methods.

(4) result and analysis

Calculate the wooden average height in area, edge of two hemisphere face image methods (THP), the area of isogonism 16 limit shape methods respectively.By expert data statistical analysis software Statistical Product and ServiceSolutions (SPSS, v13.0) (prior art), calculate the correlativity between each method, with the difference between more above-mentioned each method of paired t-test (paired t-test), the influence of whether being taken with variance analysis (ANOVA) check hemisphere face image method.The calculated in various ways forest gap area the results are shown in table 1.

Forest gap area that table 1. distinct methods records and average edge wood height

Forest gap area (the A that two hemisphere face image methods record at woods window center point _THP), shape index (SI) and edge wood average height (H _THP) be respectively 248.6 ± 66.5m ², 1.35 ± 0.03,16.9 ± 0.7m, isogonism 16 limit shape methods record forest gap area (A _ESM) be 220.1 ± 55.3m ², the edge wood average height (H) of actual measurement is 17.4 ± 0.4m (seeing Table 1).

A _THPAnd H _THPWith its detected value (A _ESMAnd H) have utmost point significant correlation respectively, related coefficient is respectively R ²=0.991 (p＜0.001), R ²=0.615 (p＜0.05).Paired t-test shows A _THPAnd H _THPAnd do not have significant difference between its detected value, testing result is respectively t=2.07 (p＞0.05) and t=0.874 (p＞0.05).A _THPCompare A _ESMBig by 13.2 ± 5.1%, H _THPLittler by-2.8 ± 3.4% than H.Variance analysis (ANOVA) shows, forest gap area (the F=0.165 that hemisphere face image method diverse location in the woods window records, p＞0.05), woods window shape index SI (F=3.205, p＞0.05), edge wood average height (F=0.184, p＞0.05) all do not have difference, promptly the hemisphere face image method is not subjected to the influence of camera site.

Comprehensive above the analysis, obtain to draw a conclusion: two hemisphere face image methods are a kind of objective, accurate and easy-operating measurement three-dimensional structure of forest gap methods of (comprising woods window size, shape and different azimuth woods window edge wood height), and there is not remarkable influence in hemisphere face filming image position to measurement result; The hemisphere face image of the woods window of taking by the time series fixed position, this method can be monitored the variation of three-dimensional structure of forest gap well, therefore, this method for woods window feature chronically, comparative study is significant.

Claims (7)

1. method of measuring three-dimensional structure of forest gap is characterized in that: the differing heights of any one position is taken two hemisphere face images straight up with the fish-eye digital camera of assembling in the woods window; Measure woods window parameter; Two hemisphere face images are handled the coordinate of record photograph middle forest window crown canopy marginal point; According to the height of fish-eye projection theory with area, shape and each orientation woods window crown canopy of each coordinate points and woods window calculation of parameter woods window;

The woods window is approximated to n limit shape, and each orientation crown canopy marginal point is apart from the height (H on ground _{T, α}) calculate by following formula:

H _t，α＝H _α+D _α·cos(α-β)·tan(γ)

Wherein, α, β, γ are respectively position angle, aspect, the gradient arbitrarily, H _α, D _αBe respectively α orientation crown canopy marginal point above camera lens height and to the vertical range of fish eye lens optical axis;

Height (the H of α orientation crown canopy marginal point above camera lens _α) and to the vertical range (D of fish eye lens optical axis _α) computing method are as follows:

H _α＝d·tan(θ _h，α)/(tan(θ _h，α)-tan(θ _l，α))+h

D _α＝d·tan(θ _h，α)·tan(θ _l，α)/(tan(θ _h，α)-tan(θ _l，α))

In the formula, d, h be respectively the discrepancy in elevation of camera lens when taking for twice, camera lens height overhead when lower is taken, θ _{L, α}, θ _{H, α}α orientation crown canopy marginal point is to the zenith angle of fish eye lens vertical optical axis when being respectively the shooting of lower and eminence;

Described zenith angle obtains by the following method:

In two hemisphere face images, write down the coordinate of each orientation crown canopy marginal point respectively, according to coordinate figure calculate the azimuth angle alpha of each orientation crown canopy marginal point and subpoint thereof to photo center point apart from r, ask the zenith angle θ of each orientation crown canopy marginal point according to fish-eye polar coordinates projection theory formula to the fish eye lens vertical optical axis:

θ/90＝r/R

In the formula, R is the radius distance of θ=90 o'clock subpoint in the hemisphere face image.

2. according to the method for the described measurement three-dimensional structure of forest gap of claim 1, it is characterized in that: the photographing request of hemisphere face image is: select optional position in the woods window of calm, full cloudy day, to be equipped with fish-eye digital camera and be placed on the tripod, take 2 hemisphere face images straight up at differing heights; Camera aperture transfers to minimum during shooting, and focal length transfers to the infinite distance, guarantee the hemisphere face image directly over for the compass direct north.

3. according to the method for the described measurement three-dimensional structure of forest gap of claim 1, it is characterized in that: adopt image processing software to handle the hemisphere face image, with the central point of hemisphere face image is that the center is since 0 degree position angle and write down the coordinate of crown canopy marginal point in the photograph at interval with fixed angle, every sheet photo writes down n=360/g coordinate points altogether, and wherein g is a fixed angle.

4. according to the method for the described measurement three-dimensional structure of forest gap of claim 1, it is characterized in that: described woods window parameter comprises the gradient and aspect.

5. according to the method for the described measurement three-dimensional structure of forest gap of claim 1, it is characterized in that: the woods window is approximated to n limit shape, calculates forest gap area (A) by following formula:

Wherein, ig is the position angle of i coordinate points, D _IgBe the vertical range of ig orientation crown canopy marginal point to the fish eye lens optical axis.

6. according to the method for the described measurement three-dimensional structure of forest gap of claim 1, it is characterized in that: the woods window is approximated to n limit shape, and the woods window shape is calculated by shape index (SI):

P is a woods window girth in the formula, and A is a forest gap area.

7. according to the method for the described measurement three-dimensional structure of forest gap of claim 6, it is characterized in that: the computing method of described woods window girth (P) are as follows:

Wherein, ig is the position angle of i coordinate points, D _IgBe the vertical range of ig orientation crown canopy marginal point to the fish eye lens optical axis.

Images